id
int64 953M
3.35B
| number
int64 2.72k
7.75k
| title
stringlengths 1
290
| state
stringclasses 2
values | created_at
timestamp[s]date 2021-07-26 12:21:17
2025-08-23 00:18:43
| updated_at
timestamp[s]date 2021-07-26 13:27:59
2025-08-23 12:34:39
| closed_at
timestamp[s]date 2021-07-26 13:27:59
2025-08-20 16:35:55
⌀ | html_url
stringlengths 49
51
| pull_request
dict | user_login
stringlengths 3
26
| is_pull_request
bool 2
classes | comments
listlengths 0
30
|
|---|---|---|---|---|---|---|---|---|---|---|---|
1,497,643,744
| 5,362
|
Run 'GPT-J' failure due to download dataset fail (' ConnectionError: Couldn't reach http://eaidata.bmk.sh/data/enron_emails.jsonl.zst ' )
|
closed
| 2022-12-15T01:23:03
| 2022-12-15T07:45:54
| 2022-12-15T07:45:53
|
https://github.com/huggingface/datasets/issues/5362
| null |
shaoyuta
| false
|
[
"Thanks for reporting, @shaoyuta.\r\n\r\nWe have checked and yes, apparently there is an issue with the server hosting the data of the \"enron_emails\" subset of \"the_pile\" dataset: http://eaidata.bmk.sh/data/enron_emails.jsonl.zst\r\nIt seems to be down: The connection has timed out.\r\n\r\nPlease note that at the Hugging Face Hub, we are not hosting their data for this dataset, but only a script that downloads the data from their servers. We are updating the data URL to one in another server.\r\n\r\nIn the meantime, please note that you can train your model in the entire \"the_pile\" dataset, by passing the \"all\" config (instead of the \"enron_emails\" one).",
"We have transferred this issue to the corresponding dataset Community tab: https://huggingface.co/datasets/the_pile/discussions/2\r\n\r\nPlease, follow the updates there."
] |
1,497,153,889
| 5,361
|
How concatenate `Audio` elements using batch mapping
|
closed
| 2022-12-14T18:13:55
| 2023-07-21T14:30:51
| 2023-07-21T14:30:51
|
https://github.com/huggingface/datasets/issues/5361
| null |
bayartsogt-ya
| false
|
[
"You can try something like this ?\r\n```python\r\ndef mapper_function(batch):\r\n return {\"concatenated_audio\": [np.concatenate([audio[\"array\"] for audio in batch[\"audio\"]])]}\r\n\r\ndataset = dataset.map(\r\n mapper_function,\r\n batched=True,\r\n batch_size=3,\r\n remove_columns=list(dataset.features),\r\n)\r\n```",
"Thanks for the snippet!\r\n\r\nOne more question. I wonder why those two mappers are working so different that one taking 4 sec while other taking over 1 min :\r\n\r\n```python\r\n%%time\r\ndef mapper_function1(batch):\r\n # list_audio\r\n return {\r\n \"audio\": [\r\n {\r\n \"array\": np.concatenate([audio[\"array\"] for audio in batch[\"audio\"]]),\r\n \"sampling_rate\": 16_000,\r\n }\r\n ]\r\n }\r\n\r\ndataset.map(\r\n mapper_function1,\r\n batched=True,\r\n batch_size=3,\r\n remove_columns=list(dataset.features),\r\n)\r\n\r\n# 100%\r\n# 135/135 [01:13<00:00, 1.93ba/s]\r\n# CPU times: user 1min 10s, sys: 3.21 s, total: 1min 13s\r\n# Wall time: 1min 13s\r\n# Dataset({\r\n# features: ['audio'],\r\n# num_rows: 135\r\n# })\r\n\r\n# --------------------------------\r\n%%time\r\ndef mapper_function2(batch):\r\n # list_audio\r\n return {\"audio\": [np.concatenate([audio[\"array\"] for audio in batch[\"audio\"]])]}\r\n\r\ndataset.map(\r\n mapper_function2,\r\n batched=True,\r\n batch_size=3,\r\n remove_columns=list(dataset.features),\r\n)\r\n\r\n# 100%\r\n# 135/135 [00:03<00:00, 40.69ba/s]\r\n# CPU times: user 1.88 s, sys: 1.48 s, total: 3.36 s\r\n# Wall time: 4.8 s\r\n# Dataset({\r\n# features: ['audio'],\r\n# num_rows: 135\r\n# })\r\n```\r\n",
"In the first one you get a dataset with an Audio type, and in the second one you get a dataset with a sequence of floats type.\r\n\r\nThe Audio type encodes the data as WAV to save disk space, so it takes more time to create.\r\nThe Audio type is automatically inferred because you modify the column \"audio\" which was already an Audio type. If you name it to something else, type inference will use a type struct with array and sampling rate fields."
] |
1,496,947,177
| 5,360
|
IterableDataset returns duplicated data using PyTorch DDP
|
closed
| 2022-12-14T16:06:19
| 2023-06-15T09:51:13
| 2023-01-16T13:33:33
|
https://github.com/huggingface/datasets/issues/5360
| null |
lhoestq
| false
|
[
"If you use huggingface trainer, you will find the trainer has wrapped a `IterableDatasetShard` to avoid duplication.\r\nSee:\r\nhttps://github.com/huggingface/transformers/blob/dfd818420dcbad68e05a502495cf666d338b2bfb/src/transformers/trainer.py#L835\r\n",
"If you want to support it by datasets natively, maybe we also need to change the code in `transformers` ?",
"Opened https://github.com/huggingface/transformers/issues/20770 to discuss this :)",
"Maybe something like this then ?\r\n```python\r\nfrom datasets.distributed import split_dataset_by_node\r\nds = split_dataset_by_node(ds, rank=rank, world_size=world_size)\r\n```\r\n\r\nFor map-style datasets the implementation is trivial (it can simply use `.shard()`).\r\n\r\nFor iterable datasets we would need to implement a new ExamplesIterable that would only iterate on a subset of the (possibly shuffled and re-shuffled after each epoch) list of shards, based on the rank and world size.",
"My plan is to skip examples by default to not end up with duplicates.\r\n\r\nAnd if a dataset has a number of shards that is a factor of the world size, then I'd make it more optimized by distributing the shards evenly across nodes instead.",
"Opened a PR here: https://github.com/huggingface/datasets/pull/5369\r\n\r\nfeel free to play with it and share your feedbacks :)",
"@lhoestq I add shuffle after split_dataset_by_node, duplicated data still exist. \r\nFor example, we have a directory named `mock_pretraining_data`, which has three files, `part-00000`, `part-00002`,`part-00002`. \r\nText in `part-00000` is like this: \r\n{\"id\": 0}\r\n{\"id\": 1}\r\n{\"id\": 2}\r\n{\"id\": 3}\r\n{\"id\": 4}\r\n{\"id\": 5}\r\n{\"id\": 6}\r\n{\"id\": 7}\r\n{\"id\": 8}\r\n{\"id\": 9}\r\n\r\nand `part-00001`\r\n{\"id\": 10}\r\n{\"id\": 11}\r\n{\"id\": 12}\r\n{\"id\": 13}\r\n{\"id\": 14}\r\n{\"id\": 15}\r\n{\"id\": 16}\r\n{\"id\": 17}\r\n{\"id\": 18}\r\n{\"id\": 19}\r\n\r\nand `part-00002`\r\n{\"id\": 20}\r\n{\"id\": 21}\r\n{\"id\": 22}\r\n{\"id\": 23}\r\n{\"id\": 24}\r\n{\"id\": 25}\r\n{\"id\": 26}\r\n{\"id\": 27}\r\n{\"id\": 28}\r\n{\"id\": 29}\r\n\r\nAnd code in `test_dist.py` like this,\r\n```python\r\nimport torch\r\nfrom torch.utils.data import Dataset, DataLoader\r\nfrom datasets import load_dataset\r\nimport os\r\nfrom transformers import AutoTokenizer, NezhaForPreTraining\r\nfrom transformers import AdamW, get_linear_schedule_with_warmup\r\nimport torch.nn.functional as F\r\nimport torch.nn as nn\r\nimport torch.distributed as dist\r\nfrom datasets.distributed import split_dataset_by_node\r\nfrom torch.nn.parallel import DistributedDataParallel as DDP\r\n\r\nos.environ[\"CUDA_VISIBLE_DEVICES\"] = '5,6,7'\r\n\r\ndist.init_process_group(\"nccl\")\r\nlocal_rank = int(os.environ['LOCAL_RANK'])\r\nworld_size = torch.distributed.get_world_size()\r\ndevice = torch.device('cuda', local_rank)\r\ndata_dir = './'\r\n\r\ndef load_trainset(train_path):\r\n dataset = load_dataset('json', data_dir=os.path.join(data_dir, train_path), split='train', streaming=True)\r\n return dataset\r\n\r\ndef collate_fn(examples):\r\n input_ids = []\r\n for example in examples:\r\n input_ids.append(example['id'])\r\n return torch.LongTensor(input_ids).to(device)\r\n\r\n\r\ndataset = load_trainset('mock_pretraining_data')\r\ndataset = split_dataset_by_node(dataset, rank=local_rank, world_size=world_size).shuffle(buffer_size=512)\r\n# train_sampler = torch.utils.data.distributed.DistributedSampler(dataset)\r\nbatch_size = 3\r\nprint('batch_size: {}'.format(batch_size))\r\ntrain_dataloader = DataLoader(dataset, batch_size=batch_size, collate_fn=collate_fn)\r\n\r\nfor x in train_dataloader:\r\n print({'rank': local_rank, 'id': x})\r\n```\r\nrun `python -m torch.distributed.launch --nproc_per_node=3 test_dist.py`\r\nThe output is\r\n```\r\n{'rank': 1, 'id': tensor([12, 15, 14], device='cuda:1')}\r\n{'rank': 1, 'id': tensor([16, 10, 18], device='cuda:1')}\r\n{'rank': 1, 'id': tensor([17, 13, 19], device='cuda:1')}\r\n{'rank': 1, 'id': tensor([11], device='cuda:1')}\r\n{'rank': 0, 'id': tensor([0, 2, 9], device='cuda:0')}\r\n{'rank': 0, 'id': tensor([4, 8, 1], device='cuda:0')}\r\n{'rank': 0, 'id': tensor([5, 3, 6], device='cuda:0')}\r\n{'rank': 0, 'id': tensor([7], device='cuda:0')}\r\n{'rank': 2, 'id': tensor([13, 15, 14], device='cuda:2')}\r\n{'rank': 2, 'id': tensor([19, 17, 18], device='cuda:2')}\r\n{'rank': 2, 'id': tensor([12, 16, 11], device='cuda:2')}\r\n{'rank': 2, 'id': tensor([10], device='cuda:2')}\r\n```\r\n`part-00001` is loaded twice, `part-00002` isn't loaded.\r\n\r\nIf I run `python -m torch.distributed.launch --nproc_per_node=2 test_dist.py`\r\nThe output is weirder，many numbers appear twice\r\n```\r\n{'rank': 1, 'id': tensor([26, 8, 13], device='cuda:1')}\r\n{'rank': 1, 'id': tensor([22, 19, 20], device='cuda:1')}\r\n{'rank': 1, 'id': tensor([12, 28, 11], device='cuda:1')}\r\n{'rank': 1, 'id': tensor([24, 2, 14], device='cuda:1')}\r\n{'rank': 1, 'id': tensor([ 6, 27, 3], device='cuda:1')}\r\n{'rank': 0, 'id': tensor([ 8, 25, 1], device='cuda:0')}\r\n{'rank': 0, 'id': tensor([20, 4, 12], device='cuda:0')}\r\n{'rank': 0, 'id': tensor([14, 29, 5], device='cuda:0')}\r\n{'rank': 0, 'id': tensor([ 7, 18, 23], device='cuda:0')}\r\n{'rank': 0, 'id': tensor([19, 17, 11], device='cuda:0')}\r\n``` ",
"Hi ! Thanks for reporting, you need to pass `seed=` to `shuffle()` or the processes won't use the same seed to shuffle the shards order before assigning each shard to a node.\r\n\r\nThe issue is that the workers are not using the same seed to shuffle the shards before splitting the shards list by node.",
"Opened https://github.com/huggingface/datasets/issues/5696",
"I have the same issue\r\n```\r\nds['train'] = load_dataset(streaming=True)\r\nds['train'] = split_dataset_by_node(ds['train'], rank=int(os.environ[\"RANK\"]), world_size=int(os.environ[\"WORLD_SIZE\"]))\r\nvectorized_datasets = ds.map(\r\n prepare_dataset,\r\n remove_columns=raw_datasets_features,\r\n).with_format(\"torch\")\r\n\r\nvectorized_datasets[\"train\"] = vectorized_datasets[\"train\"].shuffle(\r\n buffer_size=500,\r\n seed=42,\r\n)\r\n\r\ndef prepare_dataset(batch):\r\n ....\r\n print(f\"sentence: {batch['sentence']}, target_text: {batch['target_text']}\")\r\n return batch\r\n```\r\nWhen using split_dataset_by_node(), the data being read is indeed different for each GPU ID.\r\n\r\n```\r\ntrainer = Trainer(\r\n model=model,\r\n data_collator=data_collator,\r\n args=training_args,\r\n compute_metrics=compute_metrics,\r\n train_dataset=vectorized_datasets[\"train\"] if training_args.do_train else None,\r\n eval_dataset=vectorized_datasets[\"eval\"] if training_args.do_eval else None,\r\n tokenizer=processor,\r\n callbacks=[ShuffleCallback()],\r\n )\r\n...\r\ntrain_result = trainer.train(resume_from_checkpoint=checkpoint)\r\n```\r\nHowever, when I execute trainer.train(), the data being read is different from what I expected.\r\nBecause I print the batch value in prepare_dataset() , I observe that the data is the same for each GPU ID.\r\n\r\nHow should I handle this issue?\r\n\r\n\r\n",
"There are two ways an iterable dataset can be split by node:\r\n1. if the number of shards is a factor of number of GPUs: in that case the shards are evenly distributed per GPU\r\n2. otherwise, each GPU iterate on the data and at the end keeps 1 sample out of n(GPUs) - skipping the others.\r\n\r\nIn case 2. it's therefore possible to have the same examples passed to `prepare_dataset` for each GPU.\r\n\r\nThis doesn't sound optimized though, because it runs the preprocessing on samples that won't be used in the end.\r\n\r\nCould you open a new issue so that we can discuss about this and find a solution ?"
] |
1,495,297,857
| 5,359
|
Raise error if ClassLabel names is not python list
|
closed
| 2022-12-13T23:04:06
| 2022-12-22T16:35:49
| 2022-12-22T16:32:49
|
https://github.com/huggingface/datasets/pull/5359
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5359",
"html_url": "https://github.com/huggingface/datasets/pull/5359",
"diff_url": "https://github.com/huggingface/datasets/pull/5359.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5359.patch",
"merged_at": "2022-12-22T16:32:49"
}
|
freddyheppell
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"Thanks for your proposed fix, @freddyheppell.\r\n\r\nCurrently the CI fails because in a test we pass a `tuple` instead of a `list`. I would say we should accept `tuple` as a valid input type as well...\r\n\r\nWhat about checking for `Sequence` instead?",
"Fixed that @albertvillanova, can you approve CI again please? Had some issues related to Pytorch .so files when running tests on my M1 mac, so wasn't able to test locally first. Have got them working on my desktop now though."
] |
1,495,270,822
| 5,358
|
Fix `fs.open` resource leaks
|
closed
| 2022-12-13T22:35:51
| 2023-01-05T16:46:31
| 2023-01-05T15:59:51
|
https://github.com/huggingface/datasets/pull/5358
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5358",
"html_url": "https://github.com/huggingface/datasets/pull/5358",
"diff_url": "https://github.com/huggingface/datasets/pull/5358.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5358.patch",
"merged_at": "2023-01-05T15:59:51"
}
|
tkukurin
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"@mariosasko Sorry, I didn't check tests/style after doing a merge from the Git UI last week. Thx for fixing. \r\n\r\nFYI I'm getting \"Only those with [write access](https://docs.github.com/articles/what-are-the-different-access-permissions) to this repository can merge pull requests.\" so it seems somebody else needs to merge this.",
"<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008816 / 0.011353 (-0.002536) | 0.004691 / 0.011008 (-0.006317) | 0.100039 / 0.038508 (0.061531) | 0.035422 / 0.023109 (0.012313) | 0.312600 / 0.275898 (0.036702) | 0.378684 / 0.323480 (0.055204) | 0.007593 / 0.007986 (-0.000392) | 0.005183 / 0.004328 (0.000855) | 0.078040 / 0.004250 (0.073790) | 0.041845 / 0.037052 (0.004793) | 0.325251 / 0.258489 (0.066762) | 0.363459 / 0.293841 (0.069618) | 0.038006 / 0.128546 (-0.090540) | 0.011911 / 0.075646 (-0.063735) | 0.335020 / 0.419271 (-0.084251) | 0.048765 / 0.043533 (0.005233) | 0.305913 / 0.255139 (0.050774) | 0.337620 / 0.283200 (0.054420) | 0.101867 / 0.141683 (-0.039816) | 1.450091 / 1.452155 (-0.002064) | 1.437303 / 1.492716 (-0.055413) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.225650 / 0.018006 (0.207644) | 0.492480 / 0.000490 (0.491990) | 0.002857 / 0.000200 (0.002658) | 0.000075 / 0.000054 (0.000020) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.026231 / 0.037411 (-0.011180) | 0.105479 / 0.014526 (0.090953) | 0.118438 / 0.176557 (-0.058119) | 0.167313 / 0.737135 (-0.569822) | 0.119416 / 0.296338 (-0.176923) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.396233 / 0.215209 (0.181024) | 3.943325 / 2.077655 (1.865671) | 1.778864 / 1.504120 (0.274744) | 1.587957 / 1.541195 (0.046763) | 1.615404 / 1.468490 (0.146914) | 0.709427 / 4.584777 (-3.875350) | 3.823310 / 3.745712 (0.077598) | 3.461376 / 5.269862 (-1.808486) | 1.888330 / 4.565676 (-2.677346) | 0.086910 / 0.424275 (-0.337365) | 0.012215 / 0.007607 (0.004608) | 0.504877 / 0.226044 (0.278833) | 5.051513 / 2.268929 (2.782584) | 2.249389 / 55.444624 (-53.195235) | 1.890949 / 6.876477 (-4.985528) | 2.015584 / 2.142072 (-0.126489) | 0.862313 / 4.805227 (-3.942914) | 0.166295 / 6.500664 (-6.334369) | 0.061131 / 0.075469 (-0.014338) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.201804 / 1.841788 (-0.639984) | 14.589425 / 8.074308 (6.515117) | 13.855522 / 10.191392 (3.664130) | 0.193406 / 0.680424 (-0.487018) | 0.028614 / 0.534201 (-0.505587) | 0.439857 / 0.579283 (-0.139426) | 0.443330 / 0.434364 (0.008966) | 0.514078 / 0.540337 (-0.026259) | 0.608245 / 1.386936 (-0.778691) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.007087 / 0.011353 (-0.004265) | 0.005024 / 0.011008 (-0.005985) | 0.096852 / 0.038508 (0.058344) | 0.032870 / 0.023109 (0.009761) | 0.397790 / 0.275898 (0.121892) | 0.420717 / 0.323480 (0.097237) | 0.005552 / 0.007986 (-0.002434) | 0.003742 / 0.004328 (-0.000586) | 0.074788 / 0.004250 (0.070537) | 0.048030 / 0.037052 (0.010977) | 0.398520 / 0.258489 (0.140031) | 0.460919 / 0.293841 (0.167078) | 0.037652 / 0.128546 (-0.090894) | 0.012249 / 0.075646 (-0.063397) | 0.333077 / 0.419271 (-0.086194) | 0.052364 / 0.043533 (0.008831) | 0.394358 / 0.255139 (0.139219) | 0.414193 / 0.283200 (0.130994) | 0.103569 / 0.141683 (-0.038114) | 1.499208 / 1.452155 (0.047053) | 1.619481 / 1.492716 (0.126764) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.229476 / 0.018006 (0.211470) | 0.448670 / 0.000490 (0.448180) | 0.000399 / 0.000200 (0.000199) | 0.000056 / 0.000054 (0.000001) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.027550 / 0.037411 (-0.009862) | 0.109180 / 0.014526 (0.094654) | 0.118372 / 0.176557 (-0.058185) | 0.153136 / 0.737135 (-0.583999) | 0.122689 / 0.296338 (-0.173650) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.445163 / 0.215209 (0.229954) | 4.426350 / 2.077655 (2.348695) | 2.194902 / 1.504120 (0.690782) | 2.019049 / 1.541195 (0.477854) | 2.032795 / 1.468490 (0.564305) | 0.700752 / 4.584777 (-3.884025) | 3.797616 / 3.745712 (0.051903) | 2.046414 / 5.269862 (-3.223447) | 1.345037 / 4.565676 (-3.220639) | 0.085389 / 0.424275 (-0.338886) | 0.012824 / 0.007607 (0.005217) | 0.553875 / 0.226044 (0.327831) | 5.550252 / 2.268929 (3.281323) | 2.702822 / 55.444624 (-52.741803) | 2.346257 / 6.876477 (-4.530220) | 2.410772 / 2.142072 (0.268699) | 0.848271 / 4.805227 (-3.956957) | 0.170787 / 6.500664 (-6.329877) | 0.064344 / 0.075469 (-0.011125) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.266222 / 1.841788 (-0.575566) | 14.501194 / 8.074308 (6.426886) | 13.413678 / 10.191392 (3.222286) | 0.589048 / 0.680424 (-0.091375) | 0.018246 / 0.534201 (-0.515955) | 0.425221 / 0.579283 (-0.154062) | 0.425900 / 0.434364 (-0.008464) | 0.494023 / 0.540337 (-0.046314) | 0.604324 / 1.386936 (-0.782612) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png \"CML watermark\")\n"
] |
1,495,029,602
| 5,357
|
Support torch dataloader without torch formatting
|
closed
| 2022-12-13T19:39:24
| 2023-01-04T12:45:40
| 2022-12-15T19:15:54
|
https://github.com/huggingface/datasets/pull/5357
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5357",
"html_url": "https://github.com/huggingface/datasets/pull/5357",
"diff_url": "https://github.com/huggingface/datasets/pull/5357.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5357.patch",
"merged_at": "2022-12-15T19:15:54"
}
|
lhoestq
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"Need some more time to fix the tests, especially with pickle",
"> And I actually don't quite understand the idea - what's the motivation behind making only IterableDataset compatible with torch DataLoader without setting the format explicitly?\r\n\r\nSetting the format to pytorch = set the output types of the dataset to be pytorch tensors. However sometimes your dataset is not made of tensors but you still want to be able to use a pytorch DataLoader",
"A bit more context. \r\n\r\nThe arrow-backed `Dataset` supports `DataLoader(ds)` (even if the format is not \"torch\"), and we want to be able to do the same with `IterableDataset` for consistency. However, this is when the PyTorch internals come into play - an iterable dataset needs to be an instance of `torch.utils.data.IterableDataset` due to [this](https://github.com/pytorch/pytorch/blob/abc54f93145830b502400faa92bec86e05422fbd/torch/utils/data/dataloader.py#L276) check (notice there is no check for the map-style version). Hence the explicit subclassing in this PR.",
"Exactly :) Btw I just took your comments into account @polinaeterna , so feel free to review again",
"@lhoestq just checking, does this change still preserve the fix to the \"data duplicate when setting num_works > 1 with streaming data\" issue from before?\r\n\r\nhttps://github.com/huggingface/datasets/issues/3423",
"Yes :)"
] |
1,494,961,609
| 5,356
|
Clean filesystem and logging docstrings
|
closed
| 2022-12-13T18:54:09
| 2022-12-14T17:25:58
| 2022-12-14T17:22:16
|
https://github.com/huggingface/datasets/pull/5356
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5356",
"html_url": "https://github.com/huggingface/datasets/pull/5356",
"diff_url": "https://github.com/huggingface/datasets/pull/5356.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5356.patch",
"merged_at": "2022-12-14T17:22:16"
}
|
stevhliu
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,493,076,860
| 5,355
|
Clean up Table class docstrings
|
closed
| 2022-12-13T00:29:47
| 2022-12-13T18:17:56
| 2022-12-13T18:14:42
|
https://github.com/huggingface/datasets/pull/5355
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5355",
"html_url": "https://github.com/huggingface/datasets/pull/5355",
"diff_url": "https://github.com/huggingface/datasets/pull/5355.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5355.patch",
"merged_at": "2022-12-13T18:14:42"
}
|
stevhliu
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,492,174,125
| 5,354
|
Consider using "Sequence" instead of "List"
|
open
| 2022-12-12T15:39:45
| 2025-06-21T13:56:58
| null |
https://github.com/huggingface/datasets/issues/5354
| null |
tranhd95
| false
|
[
"Hi! Linking a comment to provide more info on the issue: https://stackoverflow.com/a/39458225. This means we should replace all (most of) the occurrences of `List` with `Sequence` in function signatures.\r\n\r\n@tranhd95 Would you be interested in submitting a PR?",
"Hi all! I tried to reproduce this issue and didn't work for me. Also in your example i noticed that the variables have different names: `list_of_filenames` and `list_of_files`, could this be related to that?\r\n```python\r\n#I found random data in parquet format:\r\n!wget \"https://github.com/Teradata/kylo/raw/master/samples/sample-data/parquet/userdata1.parquet\"\r\n!wget \"https://github.com/Teradata/kylo/raw/master/samples/sample-data/parquet/userdata2.parquet\"\r\n\r\n#Then i try reproduce\r\nlist_of_files = [\"userdata1.parquet\", \"userdata2.parquet\"]\r\nds = Dataset.from_parquet(list_of_files)\r\n```\r\n**My output:**\r\n```python\r\nWARNING:datasets.builder:Using custom data configuration default-e287d097dc54e046\r\nDownloading and preparing dataset parquet/default to /root/.cache/huggingface/datasets/parquet/default-e287d097dc54e046/0.0.0/2a3b91fbd88a2c90d1dbbb32b460cf621d31bd5b05b934492fdef7d8d6f236ec...\r\nDownloading data files: 100%\r\n1/1 [00:00<00:00, 40.38it/s]\r\nExtracting data files: 100%\r\n1/1 [00:00<00:00, 23.43it/s]\r\nDataset parquet downloaded and prepared to /root/.cache/huggingface/datasets/parquet/default-e287d097dc54e046/0.0.0/2a3b91fbd88a2c90d1dbbb32b460cf621d31bd5b05b934492fdef7d8d6f236ec. Subsequent calls will reuse this data.\r\n```\r\nP.S. This is my first experience with open source. So do not judge strictly if I do not understand something)",
"@dantema There is indeed a typo in variable names. Nevertheless, I'm sorry if I was not clear but the output is from `mypy` type checker. You can run the code snippet without issues. The problem is with the type checking.",
"However, I found out that the type annotation is actually misleading. The [`from_parquet`](https://github.com/huggingface/datasets/blob/5ef1ab1cc06c2b7a574bf2df454cd9fcb071ccb2/src/datasets/arrow_dataset.py#L1039) method should also accept list of [`PathLike`](https://github.com/huggingface/datasets/blob/main/src/datasets/utils/typing.py#L8) objects which includes [`os.PathLike`](https://docs.python.org/3/library/os.html#os.PathLike). But if I would ran the code snippet below, an exception is thrown.\r\n\r\n**Code**\r\n```py\r\nfrom pathlib import Path\r\n\r\nlist_of_filenames = [Path(\"foo.parquet\"), Path(\"bar.parquet\")]\r\nds = Dataset.from_parquet(list_of_filenames)\r\n```\r\n**Output**\r\n```py\r\n[/usr/local/lib/python3.8/dist-packages/datasets/arrow_dataset.py](https://localhost:8080/#) in from_parquet(path_or_paths, split, features, cache_dir, keep_in_memory, columns, **kwargs)\r\n 1071 from .io.parquet import ParquetDatasetReader\r\n 1072 \r\n-> 1073 return ParquetDatasetReader(\r\n 1074 path_or_paths,\r\n 1075 split=split,\r\n\r\n[/usr/local/lib/python3.8/dist-packages/datasets/io/parquet.py](https://localhost:8080/#) in __init__(self, path_or_paths, split, features, cache_dir, keep_in_memory, streaming, **kwargs)\r\n 35 path_or_paths = path_or_paths if isinstance(path_or_paths, dict) else {self.split: path_or_paths}\r\n 36 hash = _PACKAGED_DATASETS_MODULES[\"parquet\"][1]\r\n---> 37 self.builder = Parquet(\r\n 38 cache_dir=cache_dir,\r\n 39 data_files=path_or_paths,\r\n\r\n[/usr/local/lib/python3.8/dist-packages/datasets/builder.py](https://localhost:8080/#) in __init__(self, cache_dir, config_name, hash, base_path, info, features, use_auth_token, repo_id, data_files, data_dir, name, **config_kwargs)\r\n 298 \r\n 299 if data_files is not None and not isinstance(data_files, DataFilesDict):\r\n--> 300 data_files = DataFilesDict.from_local_or_remote(\r\n 301 sanitize_patterns(data_files), base_path=base_path, use_auth_token=use_auth_token\r\n 302 )\r\n\r\n[/usr/local/lib/python3.8/dist-packages/datasets/data_files.py](https://localhost:8080/#) in from_local_or_remote(cls, patterns, base_path, allowed_extensions, use_auth_token)\r\n 794 for key, patterns_for_key in patterns.items():\r\n 795 out[key] = (\r\n--> 796 DataFilesList.from_local_or_remote(\r\n 797 patterns_for_key,\r\n 798 base_path=base_path,\r\n\r\n[/usr/local/lib/python3.8/dist-packages/datasets/data_files.py](https://localhost:8080/#) in from_local_or_remote(cls, patterns, base_path, allowed_extensions, use_auth_token)\r\n 762 ) -> \"DataFilesList\":\r\n 763 base_path = base_path if base_path is not None else str(Path().resolve())\r\n--> 764 data_files = resolve_patterns_locally_or_by_urls(base_path, patterns, allowed_extensions)\r\n 765 origin_metadata = _get_origin_metadata_locally_or_by_urls(data_files, use_auth_token=use_auth_token)\r\n 766 return cls(data_files, origin_metadata)\r\n\r\n[/usr/local/lib/python3.8/dist-packages/datasets/data_files.py](https://localhost:8080/#) in resolve_patterns_locally_or_by_urls(base_path, patterns, allowed_extensions)\r\n 357 data_files = []\r\n 358 for pattern in patterns:\r\n--> 359 if is_remote_url(pattern):\r\n 360 data_files.append(Url(pattern))\r\n 361 else:\r\n\r\n[/usr/local/lib/python3.8/dist-packages/datasets/utils/file_utils.py](https://localhost:8080/#) in is_remote_url(url_or_filename)\r\n 62 \r\n 63 def is_remote_url(url_or_filename: str) -> bool:\r\n---> 64 parsed = urlparse(url_or_filename)\r\n 65 return parsed.scheme in (\"http\", \"https\", \"s3\", \"gs\", \"hdfs\", \"ftp\")\r\n 66 \r\n\r\n[/usr/lib/python3.8/urllib/parse.py](https://localhost:8080/#) in urlparse(url, scheme, allow_fragments)\r\n 373 Note that we don't break the components up in smaller bits\r\n 374 (e.g. netloc is a single string) and we don't expand % escapes.\"\"\"\r\n--> 375 url, scheme, _coerce_result = _coerce_args(url, scheme)\r\n 376 splitresult = urlsplit(url, scheme, allow_fragments)\r\n 377 scheme, netloc, url, query, fragment = splitresult\r\n\r\n[/usr/lib/python3.8/urllib/parse.py](https://localhost:8080/#) in _coerce_args(*args)\r\n 125 if str_input:\r\n 126 return args + (_noop,)\r\n--> 127 return _decode_args(args) + (_encode_result,)\r\n 128 \r\n 129 # Result objects are more helpful than simple tuples\r\n\r\n[/usr/lib/python3.8/urllib/parse.py](https://localhost:8080/#) in _decode_args(args, encoding, errors)\r\n 109 def _decode_args(args, encoding=_implicit_encoding,\r\n 110 errors=_implicit_errors):\r\n--> 111 return tuple(x.decode(encoding, errors) if x else '' for x in args)\r\n 112 \r\n 113 def _coerce_args(*args):\r\n\r\n[/usr/lib/python3.8/urllib/parse.py](https://localhost:8080/#) in <genexpr>(.0)\r\n 109 def _decode_args(args, encoding=_implicit_encoding,\r\n 110 errors=_implicit_errors):\r\n--> 111 return tuple(x.decode(encoding, errors) if x else '' for x in args)\r\n 112 \r\n 113 def _coerce_args(*args):\r\n\r\nAttributeError: 'PosixPath' object has no attribute 'decode'\r\n```\r\n\r\n@mariosasko Should I create a new issue? ",
"@mariosasko I would like to take this issue up. ",
"@avinashsai Hi, I've assigned you the issue.\r\n\r\n@tranhd95 Yes, feel free to report this in a new issue.",
"@avinashsai Are you still working on this? If not I would like to give it a try.",
"@mariosasko I would like to take this issue up!",
"Hi @tranhd95 @mariosasko ,I hope you all are doing well.\r\n\r\nI am interested in this issue, is this still open and unresolved ?\r\n\r\nThanks and Regards",
"@mariosasko I would like to take this issue up.",
"Hi @mariosasko, I’d like to work on this issue. I’ll start by replacing List with Sequence in relevant function signatures and ensure type checking passes. Let me know if there are any specific files to prioritize."
] |
1,491,880,500
| 5,353
|
Support remote file systems for `Audio`
|
closed
| 2022-12-12T13:22:13
| 2022-12-12T13:37:14
| 2022-12-12T13:37:14
|
https://github.com/huggingface/datasets/issues/5353
| null |
OllieBroadhurst
| false
|
[
"Just seen https://github.com/huggingface/datasets/issues/5281"
] |
1,490,796,414
| 5,352
|
__init__() got an unexpected keyword argument 'input_size'
|
open
| 2022-12-12T02:52:03
| 2022-12-19T01:38:48
| null |
https://github.com/huggingface/datasets/issues/5352
| null |
J-shel
| false
|
[
"Hi @J-shel, thanks for reporting.\r\n\r\nI think the issue comes from your call to `load_dataset`. As first argument, you should pass:\r\n- either the name of your dataset (\"mrf\") if this is already published on the Hub\r\n- or the path to the loading script of your dataset (\"path/to/your/local/mrf.py\").",
"Hi, following your suggestion, I changed my call to load_dataset. Below is the latest:\r\nreader = load_dataset('data/mrf.py',\"default\", input_size=1024, split=split, streaming=True, keep_in_memory=None)\r\nHowever, I still got the same error.\r\nI have one question that is if I only define input_size=2048 in BUILDER_CONFIGS, may I specify input_size=1024 when loading the dataset? Cause I found that I could only specify name=\"default\" since I only define name=\"default\" in BUILDER_CONFIGS."
] |
1,490,659,504
| 5,351
|
Do we need to implement `_prepare_split`?
|
closed
| 2022-12-12T01:38:54
| 2022-12-20T18:20:57
| 2022-12-12T16:48:56
|
https://github.com/huggingface/datasets/issues/5351
| null |
jmwoloso
| false
|
[
"Hi! `DatasetBuilder` is a parent class for concrete builders: `GeneratorBasedBuilder`, `ArrowBasedBuilder` and `BeamBasedBuilder`. When writing a builder script, these classes are the ones you should inherit from. And since all of them implement `_prepare_split`, you only have to implement the three methods mentioned above.",
"Thanks so much @mariosasko for the fast response! I've been referencing [this page in the docs](https://huggingface.co/docs/datasets/v2.4.0/en/about_dataset_load) because it it pretty comprehensive in terms of what we have to do and I figured since we subclass the `BuilderConfig` the same pattern would hold, but I've also seen the page with those sub-classed builders as well, so that fills in a knowledge gap for me.",
"cc @stevhliu who may have some ideas on how to improve this part of the docs.",
"one more question for my understanding @mariosasko. the requirement of a loading script has always seemed counterintuitive to me. if i have to provide a script with every dataset, what is the point of using `datasets` if we're doing all the work of loading it, I can just do that in my code and skip the datasets integration (this of course discounts other potential benefits around metadata management, etc., my example is just simplest use case though for the sake of discussion).\r\n\r\nso i figured I would implement my own `BuilderConfig` and `DatasetBuilder` to handle that portion of it and not have to make a script. i _thought_ this would result in `datasets` (via `download_and_prepare`) then making me something that I could load using `load_dataset` moving forward.\r\n\r\nConcretely, i envisioned this pattern being possible:\r\n\r\n ```\r\nclass MyBuilderConfig(BuilderConfig):\r\n def __init__(self, name=\"my_named_dataset\", ...):\r\n super().__init__(name, ...)\r\n\r\nclass MyDatasetBuilder(GeneratorBasedBuilder):\r\n BUILDER_CONFIG_CLASS = MyBuilderConfig\r\n ....\r\n\r\nmy_builder = MyDatasetBuilder(...)\r\n\r\n# this doesn't exactly work like I thought; I don't get a dataset back, but NoneType instead\r\n# though I can see it loading the files and it generates the cache, etc.\r\nmy_dataset = my_builder.download_and_prepare()\r\n\r\n# load the dataset in the future by referencing it by name and loading from the cached arrow version\r\nnew_instance_of_my_dataset = load_dataset(\"my_named_dataset\")\r\n```\r\n\r\nI've seen references to the `save_to_disk` method which might be the next step I need in order to load it by name, in which case, that makes sense, then i just need to debug why `download_and_prepare` isn't returning me a dataset, but I feel like I still have a larger conceptual knowledge gap on how to use the library correctly.\r\n\r\nThanks again in advance!",
"> the requirement of a loading script has always seemed counterintuitive to me\r\n\r\nThis is a requirement only for datasets not stored in standard formats such as CSV, JSON, SQL, Parquet, ImageFolder, etc. \r\n\r\n> if i have to provide a script with every dataset, what is the point of using datasets if we're doing all the work of loading it, I can just do that in my code and skip the datasets integration (this of course discounts other potential benefits around metadata management, etc., my example is just simplest use case though for the sake of discussion)\r\n\r\nOur README/documentation lists the main features... \r\n\r\nOne of the main ones is that our library makes it easy to work with datasets larger than RAM (thanks to Arrow and the caching mechanism), and this is not trivial to implement.\r\n\r\nRegarding the step-by-step builder, this is the pattern:\r\n```python\r\nfrom datasets import load_dataset_builder\r\nbuilder = load_dataset_builder(\"path/to/script\") # or direct instantiation with MyDatasetBuilder(...)\r\nbuilder.download_and_prepare()\r\ndset = builder.as_dataset()\r\n```",
"ok, that makes sense. thank you @mariosasko. I realized i'd never looked on the hub at any of the files associated with any datasets. just did that now and it appears that i'll need to have a script regardless _but_ that will just contain my custom config and builder classes, so without realizing it I was already making my script, I just need to wrap that in a file that sits alongside my data (I looked at Glue and realized I was already doing what I thought didn't make sense to have to do, lol).\r\n\r\n`download_and_prepare` isn't returning me a dataset though, but I'll look into that and open another issue if I can't figure it out.",
"`download_and_prepare` downloads and prepares the arrow files. You need to call `as_dataset` on the builder to get the dataset.",
"ok, I think I was assigning the output of `builder.download_and_prepare` but it's an inplace op, so that explains the `NoneType` i was getting back. Now I'm getting:\r\n\r\n```\r\nArrowInvalid Traceback (most recent call last)\r\n<ipython-input-7-3ed50fb87c70> in <module>\r\n----> 1 ds = dataset_builder.as_dataset()\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/builder.py in as_dataset(self, split, run_post_process, ignore_verifications, in_memory)\r\n 1020 \r\n 1021 # Create a dataset for each of the given splits\r\n-> 1022 datasets = map_nested(\r\n 1023 partial(\r\n 1024 self._build_single_dataset,\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/utils/py_utils.py in map_nested(function, data_struct, dict_only, map_list, map_tuple, map_numpy, num_proc, parallel_min_length, types, disable_tqdm, desc)\r\n 442 num_proc = 1\r\n 443 if num_proc <= 1 or len(iterable) < parallel_min_length:\r\n--> 444 mapped = [\r\n 445 _single_map_nested((function, obj, types, None, True, None))\r\n 446 for obj in logging.tqdm(iterable, disable=disable_tqdm, desc=desc)\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/utils/py_utils.py in <listcomp>(.0)\r\n 443 if num_proc <= 1 or len(iterable) < parallel_min_length:\r\n 444 mapped = [\r\n--> 445 _single_map_nested((function, obj, types, None, True, None))\r\n 446 for obj in logging.tqdm(iterable, disable=disable_tqdm, desc=desc)\r\n 447 ]\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/utils/py_utils.py in _single_map_nested(args)\r\n 344 # Singleton first to spare some computation\r\n 345 if not isinstance(data_struct, dict) and not isinstance(data_struct, types):\r\n--> 346 return function(data_struct)\r\n 347 \r\n 348 # Reduce logging to keep things readable in multiprocessing with tqdm\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/builder.py in _build_single_dataset(self, split, run_post_process, ignore_verifications, in_memory)\r\n 1051 \r\n 1052 # Build base dataset\r\n-> 1053 ds = self._as_dataset(\r\n 1054 split=split,\r\n 1055 in_memory=in_memory,\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/builder.py in _as_dataset(self, split, in_memory)\r\n 1120 \"\"\"\r\n 1121 cache_dir = self._fs._strip_protocol(self._output_dir)\r\n-> 1122 dataset_kwargs = ArrowReader(cache_dir, self.info).read(\r\n 1123 name=self.name,\r\n 1124 instructions=split,\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/arrow_reader.py in read(self, name, instructions, split_infos, in_memory)\r\n 236 msg = f'Instruction \"{instructions}\" corresponds to no data!'\r\n 237 raise ValueError(msg)\r\n--> 238 return self.read_files(files=files, original_instructions=instructions, in_memory=in_memory)\r\n 239 \r\n 240 def read_files(\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/arrow_reader.py in read_files(self, files, original_instructions, in_memory)\r\n 257 \"\"\"\r\n 258 # Prepend path to filename\r\n--> 259 pa_table = self._read_files(files, in_memory=in_memory)\r\n 260 # If original_instructions is not None, convert it to a human-readable NamedSplit\r\n 261 if original_instructions is not None:\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/arrow_reader.py in _read_files(self, files, in_memory)\r\n 192 f[\"filename\"] = os.path.join(self._path, f[\"filename\"])\r\n 193 for f_dict in files:\r\n--> 194 pa_table: Table = self._get_table_from_filename(f_dict, in_memory=in_memory)\r\n 195 pa_tables.append(pa_table)\r\n 196 pa_tables = [t for t in pa_tables if len(t) > 0]\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/arrow_reader.py in _get_table_from_filename(self, filename_skip_take, in_memory)\r\n 327 filename_skip_take[\"take\"] if \"take\" in filename_skip_take else None,\r\n 328 )\r\n--> 329 table = ArrowReader.read_table(filename, in_memory=in_memory)\r\n 330 if take == -1:\r\n 331 take = len(table) - skip\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/arrow_reader.py in read_table(filename, in_memory)\r\n 348 \"\"\"\r\n 349 table_cls = InMemoryTable if in_memory else MemoryMappedTable\r\n--> 350 return table_cls.from_file(filename)\r\n 351 \r\n 352 \r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/table.py in from_file(cls, filename, replays)\r\n 1034 @classmethod\r\n 1035 def from_file(cls, filename: str, replays=None):\r\n-> 1036 table = _memory_mapped_arrow_table_from_file(filename)\r\n 1037 table = cls._apply_replays(table, replays)\r\n 1038 return cls(table, filename, replays)\r\n\r\n/databricks/python/lib/python3.8/site-packages/datasets/table.py in _memory_mapped_arrow_table_from_file(filename)\r\n 48 def _memory_mapped_arrow_table_from_file(filename: str) -> pa.Table:\r\n 49 memory_mapped_stream = pa.memory_map(filename)\r\n---> 50 opened_stream = pa.ipc.open_stream(memory_mapped_stream)\r\n 51 pa_table = opened_stream.read_all()\r\n 52 return pa_table\r\n\r\n/databricks/python/lib/python3.8/site-packages/pyarrow/ipc.py in open_stream(source)\r\n 152 reader : RecordBatchStreamReader\r\n 153 \"\"\"\r\n--> 154 return RecordBatchStreamReader(source)\r\n 155 \r\n 156 \r\n\r\n/databricks/python/lib/python3.8/site-packages/pyarrow/ipc.py in __init__(self, source)\r\n 43 \r\n 44 def __init__(self, source):\r\n---> 45 self._open(source)\r\n 46 \r\n 47 \r\n\r\n/databricks/python/lib/python3.8/site-packages/pyarrow/ipc.pxi in pyarrow.lib._RecordBatchStreamReader._open()\r\n\r\n/databricks/python/lib/python3.8/site-packages/pyarrow/error.pxi in pyarrow.lib.pyarrow_internal_check_status()\r\n\r\n/databricks/python/lib/python3.8/site-packages/pyarrow/error.pxi in pyarrow.lib.check_status()\r\n\r\nArrowInvalid: Tried reading schema message, was null or length 0\r\n```\r\n\r\n",
"looks like my arrow files are all empty @mariosasko \r\n\r\n\r\n\r\n\r\ni also see the `incomplete_info.lock` file a level up too. seems like the data isn't being persisted to disk when I call `download_and_prepare`. is there something else i need to do before then, perhaps?",
"quick update @mariosasko. i got it working! i had to downgrade to `datasets==2.4.0`. testing other versions now and will let you know the results.",
"I've tested with every version of `datasets>2.4.0` and i get the same error with all of them."
] |
1,487,559,904
| 5,350
|
Clean up Loading methods docstrings
|
closed
| 2022-12-09T22:25:30
| 2022-12-12T17:27:20
| 2022-12-12T17:24:01
|
https://github.com/huggingface/datasets/pull/5350
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5350",
"html_url": "https://github.com/huggingface/datasets/pull/5350",
"diff_url": "https://github.com/huggingface/datasets/pull/5350.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5350.patch",
"merged_at": "2022-12-12T17:24:01"
}
|
stevhliu
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,487,396,780
| 5,349
|
Clean up remaining Main Classes docstrings
|
closed
| 2022-12-09T20:17:15
| 2022-12-12T17:27:17
| 2022-12-12T17:24:13
|
https://github.com/huggingface/datasets/pull/5349
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5349",
"html_url": "https://github.com/huggingface/datasets/pull/5349",
"diff_url": "https://github.com/huggingface/datasets/pull/5349.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5349.patch",
"merged_at": "2022-12-12T17:24:13"
}
|
stevhliu
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,486,975,626
| 5,348
|
The data downloaded in the download folder of the cache does not respect `umask`
|
open
| 2022-12-09T15:46:27
| 2022-12-09T17:21:26
| null |
https://github.com/huggingface/datasets/issues/5348
| null |
SaulLu
| false
|
[
"note, that `datasets` already did some of that umask fixing in the past and also at the hub - the recent work on the hub about the same: https://github.com/huggingface/huggingface_hub/pull/1220\r\n\r\nAlso I noticed that each file has a .json counterpart and the latter always has the correct perms:\r\n\r\n```\r\n-rw------- 1 uue59kq cnw 173M Dec 9 01:37 537596e64721e2ae3d98785b91d30fda0360c196a8224e29658ad629e7303a4d\r\n-rw-rw---- 1 uue59kq cnw 101 Dec 9 01:37 537596e64721e2ae3d98785b91d30fda0360c196a8224e29658ad629e7303a4d.json\r\n```\r\n\r\nso perhaps cheating is possible and syncing the perms between the 2 will do the trick."
] |
1,486,920,261
| 5,347
|
Force soundfile to return float32 instead of the default float64
|
open
| 2022-12-09T15:10:24
| 2023-01-17T16:12:49
| null |
https://github.com/huggingface/datasets/pull/5347
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5347",
"html_url": "https://github.com/huggingface/datasets/pull/5347",
"diff_url": "https://github.com/huggingface/datasets/pull/5347.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5347.patch",
"merged_at": null
}
|
qmeeus
| true
|
[
"cc @polinaeterna",
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5347). All of your documentation changes will be reflected on that endpoint.",
"Cool ! Feel free to add a comment in the code to explain that and we can merge :)",
"I'm not sure if this is a good change since we plan to get rid of `torchaudio` in the next couple of months...",
"What do you think @polinaeterna @patrickvonplaten ? Models are usually using float32 (e.g. Wev2vec2 in `transformers`) IIRC",
"IMO we can safely assume that float32 is always good enough when using audio models in inference or training. Nevertheless there might be use cases for audio datasets in the future where float64 is needed. \r\n\r\n=> I would by default always cast to float32, but then possible allow the user to cast to float64 ",
"> I'm not sure if this is a good change since we plan to get rid of torchaudio in the next couple of months...\r\n\r\n@mariosasko I agree but who knows how long we will have to wait until we are really able to do so (https://github.com/bastibe/libsndfile-binaries/pull/17 is a draft. so as @patrickvonplaten is okay with float32, I'd merge.\r\n\r\n\r\n",
"@polinaeterna Can you comment on the linked PR to see why it's still a draft? Maybe we can help somehow to get this merged finally.\r\n\r\nI think it's weird to align `soundfile` with `torchaudio` when the latter is only used for MP3 (and prob for not much longer). "
] |
1,486,884,983
| 5,346
|
[Quick poll] Give your opinion on the future of the Hugging Face Open Source ecosystem!
|
closed
| 2022-12-09T14:48:02
| 2023-06-02T20:24:44
| 2023-01-25T19:35:40
|
https://github.com/huggingface/datasets/issues/5346
| null |
LysandreJik
| false
|
[
"As the survey is finished, can we close this issue, @LysandreJik ?",
"Yes! I'll post a public summary on the forums shortly.",
"Is the summary available? I would be interested in reading your findings."
] |
1,486,555,384
| 5,345
|
Wrong dtype for array in audio features
|
open
| 2022-12-09T11:05:11
| 2023-02-10T14:39:28
| null |
https://github.com/huggingface/datasets/issues/5345
| null |
qmeeus
| false
|
[
"After some more investigation, this is due to [this line of code](https://github.com/huggingface/datasets/blob/main/src/datasets/features/audio.py#L279). The function `sf.read(file)` should be updated to `sf.read(file, dtype=\"float32\")`\r\n\r\nIndeed, the default value in soundfile is `float64` ([see here](https://pysoundfile.readthedocs.io/en/latest/#soundfile.read)). \r\n",
"@qmeeus I agree, decoding of different audio formats should return the same dtypes indeed!\r\n\r\nBut note that here you are concatenating datasets with different sampling rates: 48000 for CommonVoice and 16000 for Voxpopuli. So you should cast them to the same sampling rate value before interleaving, for example:\r\n```\r\ncv = cv.cast_column(\"audio\", Audio(sampling_rate=16000))\r\n```\r\notherwise you would get the same error because features of the same column (\"audio\") are not the same.\r\n\r\nAlso, the error you get is unexpected. Could you please confirm that you use the latest main version of the `datasets`? We had an issue that could lead to an error like this after using `rename_column` method, but it was fixed in https://github.com/huggingface/datasets/pull/5287 ",
"Hi Polina,\r\nSorry for the late answer\r\nIt is possible that the issue was due to a bug that is now fixed. I installed an editable version of datasets from github, but I don't recall whether I had updated it at the time of the issue. My research led me to other directions so I did not follow through on the interleave datasets.\r\n"
] |
1,485,628,319
| 5,344
|
Clean up Dataset and DatasetDict
|
closed
| 2022-12-09T00:02:08
| 2022-12-13T00:56:07
| 2022-12-13T00:53:02
|
https://github.com/huggingface/datasets/pull/5344
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5344",
"html_url": "https://github.com/huggingface/datasets/pull/5344",
"diff_url": "https://github.com/huggingface/datasets/pull/5344.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5344.patch",
"merged_at": "2022-12-13T00:53:01"
}
|
stevhliu
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,485,297,823
| 5,343
|
T5 for Q&A produces truncated sentence
|
closed
| 2022-12-08T19:48:46
| 2022-12-08T19:57:17
| 2022-12-08T19:57:17
|
https://github.com/huggingface/datasets/issues/5343
| null |
junyongyou
| false
|
[] |
1,485,244,178
| 5,342
|
Emotion dataset cannot be downloaded
|
closed
| 2022-12-08T19:07:09
| 2023-02-23T19:13:19
| 2022-12-09T10:46:11
|
https://github.com/huggingface/datasets/issues/5342
| null |
cbarond
| false
|
[
"Hi @cbarond there's already an open issue at https://github.com/dair-ai/emotion_dataset/issues/5, as the data seems to be missing now, so check that issue instead 👍🏻 ",
"Thanks @cbarond for reporting and @alvarobartt for pointing to the issue we opened in the author's repo.\r\n\r\nIndeed, this issue was first raised in the \"emotion\" dataset Community tab: https://huggingface.co/datasets/emotion/discussions/3\r\n\r\nI'm closing this issue and leave the issue above for the subsequent updates.\r\n\r\nDuplicate of: https://huggingface.co/datasets/emotion/discussions/3",
"try using \"SetFit/emotion\" instead",
"> try using \"SetFit/emotion\" instead\r\n\r\nI' replaced \"emotion\" with \"SetFit/Emotion\", but the code is getting stuck at\r\n\r\n`emotions = load_dataset(\"SetFit/emotion\")`\r\n\r\nI pause execution using the debugger, and it takes me to filelock.py:226\r\n\r\n`with self._thread_lock:`\r\n\r\nDo you know a way to get past this issue?",
"thanks @honeyimholm - worked for me",
"> try using \"SetFit/emotion\" instead\r\n\r\nIt really helps a lot, thank you!",
"The dataset loading script has been fixed: https://huggingface.co/datasets/emotion/discussions/4"
] |
1,484,376,644
| 5,341
|
Remove tasks.json
|
closed
| 2022-12-08T11:04:35
| 2022-12-09T12:26:21
| 2022-12-09T12:23:20
|
https://github.com/huggingface/datasets/pull/5341
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5341",
"html_url": "https://github.com/huggingface/datasets/pull/5341",
"diff_url": "https://github.com/huggingface/datasets/pull/5341.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5341.patch",
"merged_at": "2022-12-09T12:23:20"
}
|
lhoestq
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,483,182,158
| 5,340
|
Clean up DatasetInfo and Dataset docstrings
|
closed
| 2022-12-08T00:17:53
| 2022-12-08T19:33:14
| 2022-12-08T19:30:10
|
https://github.com/huggingface/datasets/pull/5340
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5340",
"html_url": "https://github.com/huggingface/datasets/pull/5340",
"diff_url": "https://github.com/huggingface/datasets/pull/5340.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5340.patch",
"merged_at": "2022-12-08T19:30:10"
}
|
stevhliu
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,482,817,424
| 5,339
|
Add Video feature, videofolder, and video-classification task
|
closed
| 2022-12-07T20:48:34
| 2024-01-11T06:30:24
| 2023-10-11T09:13:11
|
https://github.com/huggingface/datasets/pull/5339
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5339",
"html_url": "https://github.com/huggingface/datasets/pull/5339",
"diff_url": "https://github.com/huggingface/datasets/pull/5339.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5339.patch",
"merged_at": null
}
|
nateraw
| true
|
[
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5339). All of your documentation changes will be reflected on that endpoint.",
"@lhoestq I think I need some serious help with the tests 😅...I started this locally but it got too time consuming.\n\nOne issue I remember running into is with lossless audio encoding/decoding. I started thinking of using the underlying Audio feature instead of PyAV so I didn't have to rewrite similar logic here...but assumed that would turn into a mess w/ underlying logic",
"Are you still planning to work on this?",
"I'm closing this PR. Feel free to reopen it if necessary."
] |
1,482,646,151
| 5,338
|
`map()` stops every 1000 steps
|
closed
| 2022-12-07T19:09:40
| 2025-02-14T18:10:07
| 2022-12-10T00:39:28
|
https://github.com/huggingface/datasets/issues/5338
| null |
bayartsogt-ya
| false
|
[
"Hi !\r\n\r\n> It starts using all the cores (I am not sure why because I did not pass num_proc)\r\n\r\nThe tokenizer uses Rust code that is multithreaded. And maybe the `feature_extractor` might run some things in parallel as well - but I'm not super familiar with its internals.\r\n\r\n> then progress bar stops at every 1k steps. (starts using a single core)\r\n\r\nEvery 1000 examples we flush the processed examples to disk. It is this way because Arrow is a columnar format: you must write data chunk by chunk. The processing in on hold while writing right now - maybe this can be improved in the future.",
"Hi @lhoestq \r\nThanks for the explanation! it was so helpful! Let me check why `feature_extractor` is running on multiple cpus.",
"Hey @lhoestq, any news about this flush operation ? This really really slow down most of my .map and .filter operation. I'm sometime taking 10x more time waiting for the flush than actually processing the 1k items. My current workarround is to set `writer_batch_size=len(dataset)`."
] |
1,481,692,156
| 5,337
|
Support webdataset format
|
closed
| 2022-12-07T11:32:25
| 2024-03-06T14:39:29
| 2024-03-06T14:39:28
|
https://github.com/huggingface/datasets/issues/5337
| null |
lhoestq
| false
|
[
"I like the idea of having `webdataset` as an optional dependency to ensure our loader generates web datasets the same way as the main project.",
"Webdataset is the one of the most popular dataset formats for large scale computer vision tasks. Upvote for this issue. ",
"Any updates on this?",
"We haven't had the bandwidth to implement it so far, but if someone wants to give it a shot please don't hesitate ^^",
"Done in #6391 "
] |
1,479,649,900
| 5,336
|
Set `IterableDataset.map` param `batch_size` typing as optional
|
closed
| 2022-12-06T17:08:10
| 2022-12-07T14:14:56
| 2022-12-07T14:06:27
|
https://github.com/huggingface/datasets/pull/5336
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5336",
"html_url": "https://github.com/huggingface/datasets/pull/5336",
"diff_url": "https://github.com/huggingface/datasets/pull/5336.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5336.patch",
"merged_at": "2022-12-07T14:06:27"
}
|
alvarobartt
| true
|
[
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5336). All of your documentation changes will be reflected on that endpoint.",
"Hi @mariosasko, @lhoestq I was wondering whether we should include `batched` as a `pytest.mark` param for the functions testing `IterableDataset.map` so as to ensure that the changes done in this PR work fine without breaking anything of the actual functionality.\r\n\r\nI've pushed updated tests just for one of the unit testing functions to be run as `pytest tests/test_iterable_dataset.py::test_mapped_examples_iterable -s --durations 0`, but some are still missing `batched` param, it was just to ask you whether we're supposed to do this for the rest of the functions or not, if it's a yes I'll push the commit as it's ready, but didn't want to push extra stuff that may be discarded later!\r\n\r\nThanks :hugs:",
"Thanks for the feedback @lhoestq, I agree with keeping `Optional` instead of `Union[type, None]` for now 👍🏻"
] |
1,478,890,788
| 5,335
|
Update tasks.json
|
closed
| 2022-12-06T11:37:57
| 2023-09-24T10:06:42
| 2022-12-07T12:46:03
|
https://github.com/huggingface/datasets/pull/5335
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5335",
"html_url": "https://github.com/huggingface/datasets/pull/5335",
"diff_url": "https://github.com/huggingface/datasets/pull/5335.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5335.patch",
"merged_at": null
}
|
sayakpaul
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"I think the only place where we need to add it is here https://github.com/huggingface/hub-docs/blob/main/js/src/lib/interfaces/Types.ts\r\n\r\nAnd I think we can remove tasks.json completely from this repo",
"Isn't tasks.json used anymore in this repo?",
"> I think the only place where we need to add it is here https://github.com/huggingface/hub-docs/blob/main/js/src/lib/interfaces/Types.ts\r\n> \r\n> And I think we can remove tasks.json completely from this repo\r\n\r\nWhat about the warning I mentioned in https://github.com/huggingface/datasets/issues/5255#issuecomment-1339013527? Also, the depth estimation entry is already present in https://github.com/huggingface/hub-docs/blob/main/js/src/lib/interfaces/Types.ts. ",
"The update is based on what I received in the output of the export job (c.f. https://github.com/huggingface/datasets/issues/5255#issuecomment-1339107195). \r\n\r\nEdit: Oh, are you referring to the dataset card of NYU Depth V2?",
"Yes, my suggestion was for the dataset card: you got the error message because you tried to set `depth-estimation` in `class_ids` instead of `class_categories`.",
"> What about the warning I mentioned in https://github.com/huggingface/datasets/issues/5255#issuecomment-1339013527? Also, the depth estimation entry is already present in https://github.com/huggingface/hub-docs/blob/main/js/src/lib/interfaces/Types.ts.\r\n\r\nif you place it in `task_categories` you should be good :)",
"yes i would suggest rm'ing tasks.json here for clarity",
"Closing it. ",
"It's not clear if we can remove it btw, since old versions of `evaluate` rely on it (see https://github.com/huggingface/evaluate/pull/309)\r\n\r\ncc @lvwerra ",
"Actually it can be removed without incidence in old versions of evaluate since we kept an hardcoded `known_task_ids` that is marked \"DEPRECATED\""
] |
1,477,421,927
| 5,334
|
Clean up docstrings
|
closed
| 2022-12-05T20:56:08
| 2022-12-09T01:44:25
| 2022-12-09T01:41:44
|
https://github.com/huggingface/datasets/pull/5334
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5334",
"html_url": "https://github.com/huggingface/datasets/pull/5334",
"diff_url": "https://github.com/huggingface/datasets/pull/5334.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5334.patch",
"merged_at": "2022-12-09T01:41:44"
}
|
stevhliu
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"Thanks ! Let us know if we can help :)\r\n\r\nSmall pref for having multiple PRs",
"Awesome, thanks! Sorry this one is a little big, I'll open some smaller ones next :)"
] |
1,476,890,156
| 5,333
|
fix: 🐛 pass the token to get the list of config names
|
closed
| 2022-12-05T16:06:09
| 2022-12-06T08:25:17
| 2022-12-06T08:22:49
|
https://github.com/huggingface/datasets/pull/5333
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5333",
"html_url": "https://github.com/huggingface/datasets/pull/5333",
"diff_url": "https://github.com/huggingface/datasets/pull/5333.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5333.patch",
"merged_at": "2022-12-06T08:22:49"
}
|
severo
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,476,513,072
| 5,332
|
Passing numpy array to ClassLabel names causes ValueError
|
closed
| 2022-12-05T12:59:03
| 2022-12-22T16:32:50
| 2022-12-22T16:32:50
|
https://github.com/huggingface/datasets/issues/5332
| null |
freddyheppell
| false
|
[
"Should `datasets` allow `ClassLabel` input parameter to be an `np.array` even though internally we need to cast it to a Python list? @lhoestq @mariosasko ",
"Hi! No, I don't think so. The `names` parameter is [annotated](https://github.com/huggingface/datasets/blob/582236640b9109988e5f7a16a8353696ffa09a16/src/datasets/features/features.py#L892) as `List[str]` (**NumPy arrays are not lists**), and considering that type checking is not a common practice in Python, I think we can leave the code as-is.",
"I appreciate it is the wrong type, and that type checking is not common, but I think there's a few circumstances that make it a good idea from a usability perspective.\r\n\r\nIt's quite a difficult error to debug because it comes from a utility function (so it's not immediately obvious which parameter caused it). What makes it even more difficult is the exception happens when the features instance is used to instantiate the dataset, **not** when when the wrong type is actually passed when the features is instantiated. When I was debugging the error, I didn't really consider it could be an issue with the features instance because it had instantiated fine. It's also not one of the more common exceptions caused by trying to use a non-list as a list.\r\n\r\nIt's also relatively easy to accidentally get a numpy array of class types (e.g. calling `unique()` on a pandas dataframe column). Additionally, passing in a `set` instead of the list (again, relatively easy because people may run `set(classes)` to generate uniques) causes an error when the features instance is used, albeit a slightly more obvious one.\r\n\r\nThe names list is already being processed and validated in the `__post_init__` method anyway, so it would not really be adding any complexity to check it is actually a list here too. I'm happy to contribute this change if you change your mind about whether it's worthwhile.",
"I agree that it's not easy to debug this issue, so perhaps we could add some basic type checking (e.g. `not isinstance(names, list)` -> error) to make debugging easier. Feel free to submit a PR.\r\n\r\n> Additionally, passing in a set instead of the list (again, relatively easy because people may run set(classes) to generate uniques) causes an error when the features instance is used, albeit a slightly more obvious one.\r\n\r\n`set` is an unordered structure (it's ordered in Python 3.6+, but this is CPython's implementation detail), and the order of ClassLabel `names` matters, so this doesn't require a fix.",
"What about checking for `Sequence` instead? I think users can pass a list or a tuple as well."
] |
1,473,146,738
| 5,331
|
Support for multiple configs in packaged modules via metadata yaml info
|
closed
| 2022-12-02T16:43:44
| 2023-07-24T15:49:54
| 2023-07-13T13:27:56
|
https://github.com/huggingface/datasets/pull/5331
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5331",
"html_url": "https://github.com/huggingface/datasets/pull/5331",
"diff_url": "https://github.com/huggingface/datasets/pull/5331.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5331.patch",
"merged_at": "2023-07-13T13:27:56"
}
|
polinaeterna
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"feel free to merge `main` into your PR to fix the CI :)",
"Let me see if I can fix the pattern thing ^^'",
"Hmm I think it would be easier to specify the `data_files` in the end, because having a split pattern like `{split}-...` at the root of the repository can lead to unexpected behaviors IMO, and we probably don't want to have a different behavior for `data_files` depending if it's inside a `data_dir` or not\r\n\r\nMaybe something like\r\n```yaml\r\nbuilder_config:\r\n data_dir: data_dir\r\n data_files:\r\n - split: train\r\n pattern: train-[0-9][0-9][0-9][0-9]-of-[0-9][0-9][0-9][0-9][0-9]*.*\r\n```",
" > Also, I'm not sure if it's a good idea to have this field in the YAML metadata - Transformers use this part of the card only for Hub-related stuff (widgets, tags, CO2 emission, etc.), and I think we should aim to do the same in Datasets. We could achieve this by having these kwargs in a special file (they can be seen as a faster way of defining a builder (builder script) that subclasses a packaged builder) and removing the dataset_info field (the only useful info there seem to be features and we can fetch those directly from a dataset script/Parquet files).\r\n\r\nSomething like `config.json`?\r\n\r\n```json\r\n{\r\n \"data_dir\": \"data\"\r\n \"data_files\": {\r\n \"train\": \"train-[0-9][0-9][0-9][0-9]-of-[0-9][0-9][0-9][0-9][0-9]*.*\"\r\n }\r\n}\r\n```\r\n\r\nwe could also support lists for several configs",
"opened https://github.com/huggingface/datasets/issues/5694",
"I opened a PR to this PR to add data_files in YAML: https://github.com/polinaeterna/datasets/pull/1\r\n\r\n```yaml\r\nbuilder_config:\r\n data_files:\r\n - split: train\r\n pattern: data/train-*\r\n```",
"Let me open a PR to see if I can move the data files resolution outside of the MetadataConfigs to not modify it in-place",
"I wonder if we can make the cache backward compatible: we could just check if the cache directory with the old path exists. It will be useful for the research team which has a big datasets cache",
"> I wonder if we can make the cache backward compatible: we could just check if the cache directory with the old path exists. It will be useful for the research team which has a big datasets cache\r\n\r\n\r\n\r\nIn the next PR maybe? :D \r\nIt's possible but requires some additional logic to correctly pass old `config_kwargs` (which used to include `data_files` but now it's `None` for builders from metadata) to generate the hash which is used to create the path.",
"If we only consider datasets that were pushed to hub, it's just a matter of using `\"{username}__parquet\"` instead of `\"{username}__{dataset_name}\"` in the cache directory name. The hashes stay the same :)\r\n\r\nEDIT: and the config name\r\nEDIT2: and the arrow file names",
"Did a small PR for backward compatibility, it was easy to add in the end: https://github.com/polinaeterna/datasets/pull/3",
"Just created a branch [dev-3.0](https://github.com/huggingface/datasets/tree/dev-3.0) in which we can merge this one and the other datasets 3.0 related PRs",
"@lhoestq why can't we merge it in main?",
"We can, it was just in case we had other things to merge after @mariosasko or @albertvillanova 's reviews",
"@lhoestq @albertvillanova @mariosasko we agreed on having `configs` (in plural) as a metadata field in readme but apparently Hub's yaml validation doesn't allow it to be not a list :D \r\n\r\n(with `config` (in singular) it works)\r\n\r\nedit: and now the tests for hub datasets with metadata configs are failing because I cannot change the yaml there...",
"> we agreed on having configs (in plural) as a metadata field in readme but apparently Hub's yaml validation doesn't allow it to be not a list :D\r\n\r\nIf the `configs` field is specified in the YAML, the Hub can use it to [improve](https://github.com/huggingface/moon-landing/blob/97aca4cac32fbb7d84ce5eba9b18afad87968c4a/server/views/components/DatasetLibraryModal/datasetLibrarySnippets.ts#L11) the `Use in dataset library` snippet by listing the possible config values in `load_dataset`. So I think this needs to be fixed on the Hub side.\r\n\r\nPS: I couldn't find an instance of someone using this field on the Hub, so I think using it for this feature is OK.",
"> I couldn't find an instance of someone using this field on the Hub, so I think using it for this feature is OK.\r\n\r\n@mariosasko I think it's because @lhoestq renamed `configs` to `config_names` in all canonical datasets :D so yes, `configs` field is now supposed to include custom configuration parameters introduced in this PR, and `config_names` is used (not really used lol) for list of strings of config names. It's being fixed on the Hub's side https://github.com/huggingface/moon-landing/pull/6490",
"after more thought I agree it's maybe overkill to do a major release for this one, since we have a good backward compatibility",
"There is one edge case I forgot to mention in the reviews - I think it's a good idea to support passing config params that are functions (Pandas uses them a lot) using this API (e.g. `converters` in the CSV config for converting a string column into a sequence). I see two solutions: string blocks with Python code in YAML or PyYAML [tags](https://pyyaml.org/wiki/PyYAMLDocumentation#yaml-tags-and-python-types). \r\n\r\nBut I think this can be addressed later.",
"I'm resolving the conflicts and writing some docs :) let's merge this soon !",
"<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==8.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.005868 / 0.011353 (-0.005485) | 0.003544 / 0.011008 (-0.007464) | 0.080329 / 0.038508 (0.041821) | 0.061072 / 0.023109 (0.037963) | 0.307802 / 0.275898 (0.031904) | 0.340353 / 0.323480 (0.016873) | 0.004665 / 0.007986 (-0.003321) | 0.002779 / 0.004328 (-0.001550) | 0.062065 / 0.004250 (0.057815) | 0.046350 / 0.037052 (0.009297) | 0.312045 / 0.258489 (0.053556) | 0.353524 / 0.293841 (0.059683) | 0.026965 / 0.128546 (-0.101581) | 0.007906 / 0.075646 (-0.067740) | 0.260678 / 0.419271 (-0.158593) | 0.044167 / 0.043533 (0.000634) | 0.309757 / 0.255139 (0.054618) | 0.340188 / 0.283200 (0.056988) | 0.020440 / 0.141683 (-0.121243) | 1.486886 / 1.452155 (0.034732) | 1.548330 / 1.492716 (0.055614) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.188658 / 0.018006 (0.170652) | 0.422204 / 0.000490 (0.421715) | 0.003508 / 0.000200 (0.003308) | 0.000068 / 0.000054 (0.000013) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.025173 / 0.037411 (-0.012238) | 0.072868 / 0.014526 (0.058343) | 0.084817 / 0.176557 (-0.091739) | 0.151667 / 0.737135 (-0.585468) | 0.085632 / 0.296338 (-0.210706) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.400998 / 0.215209 (0.185789) | 4.022274 / 2.077655 (1.944619) | 2.025768 / 1.504120 (0.521648) | 1.874193 / 1.541195 (0.332998) | 2.006537 / 1.468490 (0.538047) | 0.501799 / 4.584777 (-4.082978) | 2.987487 / 3.745712 (-0.758225) | 4.552295 / 5.269862 (-0.717566) | 2.775859 / 4.565676 (-1.789817) | 0.057596 / 0.424275 (-0.366679) | 0.006449 / 0.007607 (-0.001158) | 0.470776 / 0.226044 (0.244732) | 4.725933 / 2.268929 (2.457005) | 2.480130 / 55.444624 (-52.964494) | 2.183919 / 6.876477 (-4.692558) | 2.408052 / 2.142072 (0.265979) | 0.584038 / 4.805227 (-4.221190) | 0.124964 / 6.500664 (-6.375701) | 0.060939 / 0.075469 (-0.014530) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.221263 / 1.841788 (-0.620524) | 18.326372 / 8.074308 (10.252064) | 13.398937 / 10.191392 (3.207545) | 0.149153 / 0.680424 (-0.531271) | 0.016941 / 0.534201 (-0.517260) | 0.332106 / 0.579283 (-0.247177) | 0.339958 / 0.434364 (-0.094406) | 0.378125 / 0.540337 (-0.162212) | 0.517787 / 1.386936 (-0.869149) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.005927 / 0.011353 (-0.005426) | 0.003607 / 0.011008 (-0.007402) | 0.062925 / 0.038508 (0.024417) | 0.058676 / 0.023109 (0.035566) | 0.362129 / 0.275898 (0.086231) | 0.395864 / 0.323480 (0.072384) | 0.004652 / 0.007986 (-0.003334) | 0.002893 / 0.004328 (-0.001435) | 0.062696 / 0.004250 (0.058445) | 0.049988 / 0.037052 (0.012935) | 0.365366 / 0.258489 (0.106877) | 0.412326 / 0.293841 (0.118485) | 0.027118 / 0.128546 (-0.101429) | 0.008179 / 0.075646 (-0.067467) | 0.068048 / 0.419271 (-0.351223) | 0.041065 / 0.043533 (-0.002468) | 0.359858 / 0.255139 (0.104719) | 0.386589 / 0.283200 (0.103390) | 0.020467 / 0.141683 (-0.121216) | 1.438070 / 1.452155 (-0.014084) | 1.479617 / 1.492716 (-0.013099) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.231516 / 0.018006 (0.213510) | 0.413407 / 0.000490 (0.412917) | 0.000358 / 0.000200 (0.000158) | 0.000052 / 0.000054 (-0.000002) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.026071 / 0.037411 (-0.011340) | 0.076486 / 0.014526 (0.061960) | 0.085943 / 0.176557 (-0.090613) | 0.138087 / 0.737135 (-0.599048) | 0.087466 / 0.296338 (-0.208872) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.417711 / 0.215209 (0.202502) | 4.171915 / 2.077655 (2.094260) | 2.140677 / 1.504120 (0.636557) | 1.960164 / 1.541195 (0.418969) | 2.002134 / 1.468490 (0.533644) | 0.499699 / 4.584777 (-4.085078) | 2.991814 / 3.745712 (-0.753898) | 2.906589 / 5.269862 (-2.363272) | 1.842305 / 4.565676 (-2.723372) | 0.057633 / 0.424275 (-0.366642) | 0.006465 / 0.007607 (-0.001142) | 0.492874 / 0.226044 (0.266830) | 4.931613 / 2.268929 (2.662684) | 2.623161 / 55.444624 (-52.821463) | 2.310624 / 6.876477 (-4.565853) | 2.483146 / 2.142072 (0.341074) | 0.586910 / 4.805227 (-4.218317) | 0.124681 / 6.500664 (-6.375983) | 0.061561 / 0.075469 (-0.013908) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.319111 / 1.841788 (-0.522677) | 18.637326 / 8.074308 (10.563018) | 13.803912 / 10.191392 (3.612520) | 0.143989 / 0.680424 (-0.536435) | 0.017025 / 0.534201 (-0.517176) | 0.333156 / 0.579283 (-0.246127) | 0.342163 / 0.434364 (-0.092201) | 0.380357 / 0.540337 (-0.159981) | 0.512261 / 1.386936 (-0.874675) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#f49a16346dc35e5eabeec39778d0f2e4e850dfd7 \"CML watermark\")\n"
] |
1,471,999,125
| 5,329
|
Clarify imagefolder is for small datasets
|
closed
| 2022-12-01T21:47:29
| 2022-12-06T17:20:04
| 2022-12-06T17:16:53
|
https://github.com/huggingface/datasets/pull/5329
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5329",
"html_url": "https://github.com/huggingface/datasets/pull/5329",
"diff_url": "https://github.com/huggingface/datasets/pull/5329.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5329.patch",
"merged_at": "2022-12-06T17:16:53"
}
|
stevhliu
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"I think it's also reasonable to add the same note to the AudioFolder decription",
"Thank you ! I think \"regular\" is more appropriate than \"small\". It can easily scale to a few thousands of images - just not millions x)",
"Replaced \"small\" with \"several thousand\" since what is considered \"regular\" and even \"small\" can be kind of vague!"
] |
1,471,661,437
| 5,328
|
Fix docs building for main
|
closed
| 2022-12-01T17:07:45
| 2022-12-02T16:29:00
| 2022-12-02T16:26:00
|
https://github.com/huggingface/datasets/pull/5328
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5328",
"html_url": "https://github.com/huggingface/datasets/pull/5328",
"diff_url": "https://github.com/huggingface/datasets/pull/5328.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5328.patch",
"merged_at": "2022-12-02T16:26:00"
}
|
albertvillanova
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"EDIT\r\nAt least the docs for ~~main~~ PR branch are now built:\r\n- https://github.com/huggingface/datasets/actions/runs/3594847760/jobs/6053620813",
"Build documentation for main branch was triggered after this PR being merged: https://github.com/huggingface/datasets/actions/runs/3603370082/jobs/6071482470"
] |
1,471,657,247
| 5,327
|
Avoid unwanted behaviour when splits from script and metadata are not matching because of outdated metadata
|
open
| 2022-12-01T17:05:23
| 2023-01-23T12:48:29
| null |
https://github.com/huggingface/datasets/pull/5327
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5327",
"html_url": "https://github.com/huggingface/datasets/pull/5327",
"diff_url": "https://github.com/huggingface/datasets/pull/5327.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5327.patch",
"merged_at": null
}
|
polinaeterna
| true
|
[
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5327). All of your documentation changes will be reflected on that endpoint."
] |
1,471,634,168
| 5,326
|
No documentation for main branch is built
|
closed
| 2022-12-01T16:50:58
| 2022-12-02T16:26:01
| 2022-12-02T16:26:01
|
https://github.com/huggingface/datasets/issues/5326
| null |
albertvillanova
| false
|
[] |
1,471,536,822
| 5,325
|
map(...batch_size=None) for IterableDataset
|
closed
| 2022-12-01T15:43:42
| 2022-12-07T15:54:43
| 2022-12-07T15:54:42
|
https://github.com/huggingface/datasets/issues/5325
| null |
frankier
| false
|
[
"Hi! I agree it makes sense for `IterableDataset.map` to support the `batch_size=None` case. This should be super easy to fix.",
"@mariosasko as this is something simple maybe I can include it as part of https://github.com/huggingface/datasets/pull/5311? Let me know :+1:",
"#self-assign",
"Feel free to close this @lhoestq as part of https://github.com/huggingface/datasets/pull/5336 :hugs:",
"Thanks again :)\r\n\r\n> For practical usages, an alternative to this would be to convert from an iterable dataset to a map-style dataset, but it is not obvious how to do this.\r\n\r\nThis is interesting as well, if anyone wants to explore"
] |
1,471,524,512
| 5,324
|
Fix docstrings and types in documentation that appears on the website
|
open
| 2022-12-01T15:34:53
| 2024-01-23T16:21:54
| null |
https://github.com/huggingface/datasets/issues/5324
| null |
polinaeterna
| false
|
[
"I agree we have a mess with docstrings...",
"Ok, I believe we've cleaned up most of the old syntax we were using for the user-facing docs! There are still a couple of `:obj:`'s and `:class:` floating around in the docstrings we don't expose that I'll track down :)",
"Hi @polinaeterna @albertvillanova @stevhliu, I hope you all are doing well.\r\n\r\nIs this issue still unresolved as I am interested in it?",
"It should be mostly fixed for the user-facing APIs, but there may be some Sphinx syntax still lurking around in the non-public APIs. Feel free to open a PR to fix those if you catch any! 🤗 ",
"Thanks for your reply @stevhliu :)\r\nSure, I will try to find out the remaining and fix that.\r\n\r\n"
] |
1,471,518,803
| 5,323
|
Duplicated Keys in Taskmaster-2 Dataset
|
closed
| 2022-12-01T15:31:06
| 2022-12-01T16:26:06
| 2022-12-01T16:26:06
|
https://github.com/huggingface/datasets/issues/5323
| null |
liaeh
| false
|
[
"Thanks for reporting, @liaeh.\r\n\r\nWe are having a look at it. ",
"I have transferred the discussion to the Community tab of the dataset: https://huggingface.co/datasets/taskmaster2/discussions/1"
] |
1,471,502,162
| 5,322
|
Raise error for `.tar` archives in the same way as for `.tar.gz` and `.tgz` in `_get_extraction_protocol`
|
closed
| 2022-12-01T15:19:28
| 2022-12-14T16:37:16
| 2022-12-14T16:33:30
|
https://github.com/huggingface/datasets/pull/5322
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5322",
"html_url": "https://github.com/huggingface/datasets/pull/5322",
"diff_url": "https://github.com/huggingface/datasets/pull/5322.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5322.patch",
"merged_at": "2022-12-14T16:33:30"
}
|
polinaeterna
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,471,430,667
| 5,321
|
Fix loading from HF GCP cache
|
closed
| 2022-12-01T14:39:06
| 2022-12-01T16:10:09
| 2022-12-01T16:07:02
|
https://github.com/huggingface/datasets/pull/5321
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5321",
"html_url": "https://github.com/huggingface/datasets/pull/5321",
"diff_url": "https://github.com/huggingface/datasets/pull/5321.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5321.patch",
"merged_at": "2022-12-01T16:07:02"
}
|
lhoestq
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"> Do you know why this stopped working?\r\n\r\nIt comes from the changes in https://github.com/huggingface/datasets/pull/5107/files#diff-355ae5c229f95f86895404b72378ecd6e966c41cbeebb674af6fe6e9611bc126"
] |
1,471,360,910
| 5,320
|
[Extract] Place the lock file next to the destination directory
|
closed
| 2022-12-01T13:55:49
| 2022-12-01T15:36:44
| 2022-12-01T15:33:58
|
https://github.com/huggingface/datasets/pull/5320
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5320",
"html_url": "https://github.com/huggingface/datasets/pull/5320",
"diff_url": "https://github.com/huggingface/datasets/pull/5320.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5320.patch",
"merged_at": "2022-12-01T15:33:58"
}
|
lhoestq
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,470,945,515
| 5,319
|
Fix Text sample_by paragraph
|
closed
| 2022-12-01T09:08:09
| 2022-12-01T15:21:44
| 2022-12-01T15:19:00
|
https://github.com/huggingface/datasets/pull/5319
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5319",
"html_url": "https://github.com/huggingface/datasets/pull/5319",
"diff_url": "https://github.com/huggingface/datasets/pull/5319.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5319.patch",
"merged_at": "2022-12-01T15:19:00"
}
|
albertvillanova
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,470,749,750
| 5,318
|
Origin/fix missing features error
|
closed
| 2022-12-01T06:18:39
| 2022-12-12T19:06:42
| 2022-12-04T05:49:39
|
https://github.com/huggingface/datasets/pull/5318
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5318",
"html_url": "https://github.com/huggingface/datasets/pull/5318",
"diff_url": "https://github.com/huggingface/datasets/pull/5318.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5318.patch",
"merged_at": "2022-12-04T05:49:39"
}
|
eunseojo
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"please review :) @lhoestq @ola13 thankoo",
"Thanks :) I just updated the test to make sure it works even when there's a column missing, and did a minor change to json.py to add the missing columns for the other kinds of JSON files as well (I moved the code to`self._cast_table`)",
"Thanks Unso! If @lhoestq is happy then I'm also happy :D",
"When I noticed the ping, this PR had already been merged...\r\n\r\nLuckily, PyArrow's `read_json` behaves the same when `explicit_schema` is given via `ParseOptions`, so I'm okay with this change (our JSON loader doesn't use `read_json` for decoding JSON in some scenarios, so this manual approach is the right one).\r\n"
] |
1,470,390,164
| 5,317
|
`ImageFolder` performs poorly with large datasets
|
open
| 2022-12-01T00:04:21
| 2022-12-01T21:49:26
| null |
https://github.com/huggingface/datasets/issues/5317
| null |
salieri
| false
|
[
"Hi ! ImageFolder is made for small scale datasets indeed. For large scale image datasets you better group your images in TAR archives or Arrow/Parquet files. This is true not just for ImageFolder loading performance, but also because having millions of files is not ideal for your filesystem or when moving the data around.\r\n\r\nOption 1. use TAR archives\r\n\r\nI'd suggest you to take a look at how we load [Imagenet](https://huggingface.co/datasets/imagenet-1k/tree/main) for example. The dataset is sharded in multiple TAR archives and there is a [script](https://huggingface.co/datasets/imagenet-1k/blob/main/imagenet-1k.py) that iterates over the archives to load the images.\r\n\r\nOption 2. use Arrow/Parquet\r\n\r\nYou can load your images as an Arrow Dataset with\r\n```python\r\nfrom datasets import Dataset, Image, load_from_disk, load_dataset\r\n\r\nds = Dataset.from_dict({\"image\": list(glob.glob(\"path/to/dir/**/*.jpg\"))})\r\n\r\ndef add_metadata(example):\r\n ...\r\n\r\nds = ds.map(add_metadata, num_proc=16) # num_proc for multiprocessing\r\nds = ds.cast_column(\"image\", Image())\r\n\r\n# save as Arrow locally\r\nds.save_to_disk(\"output_dir\")\r\nreloaded = load_from_disk(\"output_dir\")\r\n\r\n# OR save as Parquet on the HF Hub\r\nds.push_to_hub(\"username/dataset_name\")\r\nreloaded = load_dataset(\"username/dataset_name\")\r\n# reloaded = load_dataset(\"username/dataset_name\", num_proc=16) # to use multiprocessing\r\n```\r\n\r\nPS: maybe we can actually have something similar to ImageFolder but for image archives at one point ?",
"@lhoestq Thanks!\r\n\r\nPerhaps it'd be worth adding a note on the documentation that `ImageFolder` is not intended for large datasets? This limitation is not intuitively obvious to someone who has not used it before, I think.",
"Thanks for the feedback @salieri! I opened #5329 to make it clear `ImageFolder` is not intended for large datasets. Please feel free to comment if you have any other feedback! 🙂 "
] |
1,470,115,681
| 5,316
|
Bug in sample_by="paragraph"
|
closed
| 2022-11-30T19:24:13
| 2022-12-01T15:19:02
| 2022-12-01T15:19:02
|
https://github.com/huggingface/datasets/issues/5316
| null |
adampauls
| false
|
[
"Thanks for reporting, @adampauls.\r\n\r\nWe are having a look at it. "
] |
1,470,026,797
| 5,315
|
Adding new splits to a dataset script with existing old splits info in metadata's `dataset_info` fails
|
open
| 2022-11-30T18:02:15
| 2022-12-02T07:02:53
| null |
https://github.com/huggingface/datasets/issues/5315
| null |
polinaeterna
| false
|
[
"EDIT:\r\nI think in this case, the metadata files (either README or JSON) should not be read (i.e. `self.info.splits` should be None).\r\n\r\nOne idea: \r\n- I think ideally we should set this behavior when we pass `--save_info` to the CLI `test`\r\n- However, currently, the builder is unaware of this: `save_info` arg is not passed to it",
"> I think in this case\r\n\r\n@albertvillanova You mean in cases when the script was changed? \r\n\r\nI suggest that we:\r\n* add a check on the slice (like 'split_name[n%]) kind of format here: https://github.com/huggingface/datasets/blob/main/src/datasets/splits.py#L523 to catch things like this. \r\n* Error here happens before splits verification, but in `_prepare_split`, and `_prepare_split` doesn't perform any verification and don't know about it. so we can pass this parameter and take splits from `split_generator`, not from `split.info` in case when `verify_infos` is False\r\n* we can check if split **names** from split_generators and self.info.splits are the same **before** preparing splits (if `verify_info=True`) so that we don't spend time on generating unwanted data. \r\n* provide some user-friendly warnings about `ignore_verifications` parameter so that users know that if something is not matching they can ignore it\r\n\r\nI started it here: https://github.com/huggingface/datasets/pull/5327/files\r\n\r\nWhat do you think @albertvillanova ?",
"I edited my previous comment:\r\n- First I proposed setting `self.info.splits` to None when `ignore_verifications=True`\r\n - I thought it was the easiest implementation because `ignore_verifications` is passed to `DatasetBuilder.download_and_prepare`\r\n - However, afterwards, I realized this might not be a good idea for this use case:\r\n - A user wants to optimize the loading of the dataset, and passes `ignore_verifications=False` to avoid all the verifications\r\n - In this case, we want `self.info.splits` to be read from metadata file\r\n- Then, I thought that it might be better to set `self.info.splits` to None when we pass `--save_info` to the CLI test: if we are going to save the info to the metadata file, it makes no sense to read the info from the metadata file\r\n - This implementation is not so easy because the Builder knows nothing about `--save_info`\r\n\r\nI agree with you there are 2 things to be addressed here:\r\n- One is what I have just commented: `self.info.splits` should be None in this case\r\n- The other, a validation should be implemented when calling `make_file_instructions` and/or `SplitDict.__getitem__`, so that when passing \"training\" to it, we get a more descriptive error other than `TypeError: expected str, bytes or os.PathLike object, not NoneType` "
] |
1,469,685,118
| 5,314
|
Datasets: classification_report() got an unexpected keyword argument 'suffix'
|
closed
| 2022-11-30T14:01:03
| 2023-07-21T14:40:31
| 2023-07-21T14:40:31
|
https://github.com/huggingface/datasets/issues/5314
| null |
JonathanAlis
| false
|
[
"This seems similar to https://github.com/huggingface/datasets/issues/2512 Can you try to update seqeval ? ",
"@JonathanAlis also note that the metrics are deprecated in our `datasets` library.\r\n\r\nPlease, use the new library 🤗 Evaluate instead: https://huggingface.co/docs/evaluate"
] |
1,468,484,136
| 5,313
|
Fix description of streaming in the docs
|
closed
| 2022-11-29T18:00:28
| 2022-12-01T14:55:30
| 2022-12-01T14:00:34
|
https://github.com/huggingface/datasets/pull/5313
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5313",
"html_url": "https://github.com/huggingface/datasets/pull/5313",
"diff_url": "https://github.com/huggingface/datasets/pull/5313.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5313.patch",
"merged_at": "2022-12-01T14:00:34"
}
|
polinaeterna
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,468,352,562
| 5,312
|
Add DatasetDict.to_pandas
|
closed
| 2022-11-29T16:30:02
| 2023-09-24T10:06:19
| 2023-01-25T17:33:42
|
https://github.com/huggingface/datasets/pull/5312
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5312",
"html_url": "https://github.com/huggingface/datasets/pull/5312",
"diff_url": "https://github.com/huggingface/datasets/pull/5312.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5312.patch",
"merged_at": null
}
|
lhoestq
| true
|
[
"The current implementation is what I had in mind, i.e. concatenate all splits by default.\r\n\r\nHowever, I think most tabular datasets would come as a single split. So for that usecase, it wouldn't change UX if we raise when there are more than one splits.\r\n\r\nAnd for multiple splits, the user either passes a list, or they can pass `splits=\"all\"` to have all splits concatenated.",
"I think it's better to raise an error in cases when there are multiple splits but no split is specified so that users know for sure with which data they are working. I imagine a case when a user loads a dataset that they don't know much about (like what splits it has), and if they get a concatenation of everything, it might lead to incorrect processing or interpretations and it would be hard to notice it.\r\n(\"explicit is better than implicit\")",
"I just changed to raise an error if there are multiple splits. The error shows an example of how to choose a split to convert.",
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5312). All of your documentation changes will be reflected on that endpoint.",
"Thanks for the review, I've updated the type hint and added a line to raise an error on bad splits :)",
"Merging https://github.com/huggingface/datasets/pull/5301 would eliminate the need for this PR, no?\r\n\r\nIn the meantime, I find the current API cleaner.",
"This solution is simpler than https://github.com/huggingface/datasets/pull/5301 and covers most cases for tabular datasets, so I'm in favor of merging this one and put https://github.com/huggingface/datasets/pull/5301 on stand by",
"Let me know if it sounds good to you @mariosasko @albertvillanova :)",
"I'm still not convinced. If `DatasetDict` needs this method and there is no other way, then IMO it would make more sense to return a dictionary with the splits converted to `pd.DataFrame`. ",
"@mariosasko the issue we're dealing with is that in tabular scenarios, we often don't have splits in the dataset, and imposing that concept to people dealing with the library hampers adoption.",
"@adrinjalali This PR proposes a solution inconsistent with the existing API (in other words, a solution that clutters our API 🙂). Moreover, our library primarily focuses on larger-than-RAM datasets, and tabular datasets don't (directly) fall into this group.\r\n\r\nInstead of the temporary \"fix\" proposed here, it makes much more sense to align `load_dataset` with both tabular and DL workflows \"in a consistent way\", so I suggest we continue our discussion from https://github.com/huggingface/datasets/issues/5189 to have this resolved by version 3.0.",
"closing this one for now"
] |
1,467,875,153
| 5,311
|
Add `features` param to `IterableDataset.map`
|
closed
| 2022-11-29T11:08:34
| 2022-12-06T15:45:02
| 2022-12-06T15:42:04
|
https://github.com/huggingface/datasets/pull/5311
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5311",
"html_url": "https://github.com/huggingface/datasets/pull/5311",
"diff_url": "https://github.com/huggingface/datasets/pull/5311.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5311.patch",
"merged_at": "2022-12-06T15:42:04"
}
|
alvarobartt
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,467,719,635
| 5,310
|
Support xPath for Windows pathnames
|
closed
| 2022-11-29T09:20:47
| 2022-11-30T12:00:09
| 2022-11-30T11:57:16
|
https://github.com/huggingface/datasets/pull/5310
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5310",
"html_url": "https://github.com/huggingface/datasets/pull/5310",
"diff_url": "https://github.com/huggingface/datasets/pull/5310.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5310.patch",
"merged_at": "2022-11-30T11:57:16"
}
|
albertvillanova
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,466,758,987
| 5,309
|
Close stream in `ArrowWriter.finalize` before inference error
|
closed
| 2022-11-28T16:59:39
| 2022-12-07T12:55:20
| 2022-12-07T12:52:15
|
https://github.com/huggingface/datasets/pull/5309
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5309",
"html_url": "https://github.com/huggingface/datasets/pull/5309",
"diff_url": "https://github.com/huggingface/datasets/pull/5309.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5309.patch",
"merged_at": "2022-12-07T12:52:15"
}
|
mariosasko
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,466,552,281
| 5,308
|
Support `topdown` parameter in `xwalk`
|
closed
| 2022-11-28T14:42:41
| 2022-12-09T12:58:55
| 2022-12-09T12:55:59
|
https://github.com/huggingface/datasets/pull/5308
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5308",
"html_url": "https://github.com/huggingface/datasets/pull/5308",
"diff_url": "https://github.com/huggingface/datasets/pull/5308.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5308.patch",
"merged_at": "2022-12-09T12:55:59"
}
|
mariosasko
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"I like the `kwargs` approach, thanks!"
] |
1,466,477,427
| 5,307
|
Use correct dataset type in `from_generator` docs
|
closed
| 2022-11-28T13:59:10
| 2022-11-28T15:30:37
| 2022-11-28T15:27:26
|
https://github.com/huggingface/datasets/pull/5307
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5307",
"html_url": "https://github.com/huggingface/datasets/pull/5307",
"diff_url": "https://github.com/huggingface/datasets/pull/5307.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5307.patch",
"merged_at": "2022-11-28T15:27:26"
}
|
mariosasko
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,465,968,639
| 5,306
|
Can't use custom feature description when loading a dataset
|
closed
| 2022-11-28T07:55:44
| 2022-11-28T08:11:45
| 2022-11-28T08:11:44
|
https://github.com/huggingface/datasets/issues/5306
| null |
clefourrier
| false
|
[
"Forgot to actually convert the feature dict to a Feature object. Closing."
] |
1,465,627,826
| 5,305
|
Dataset joelito/mc4_legal does not work with multiple files
|
closed
| 2022-11-28T00:16:16
| 2022-11-28T07:22:42
| 2022-11-28T07:22:42
|
https://github.com/huggingface/datasets/issues/5305
| null |
JoelNiklaus
| false
|
[
"Thanks for reporting @JoelNiklaus.\r\n\r\nPlease note that since we moved all dataset loading scripts to the Hub, the issues and pull requests relative to specific datasets are directly handled on the Hub, in their Community tab. I'm transferring this issue there: https://huggingface.co/datasets/joelito/mc4_legal/discussions\r\n\r\nI am also having a look at the bug in your script.",
"Issue transferred to: https://huggingface.co/datasets/joelito/mc4_legal/discussions/1"
] |
1,465,110,367
| 5,304
|
timit_asr doesn't load the test split.
|
closed
| 2022-11-26T10:18:22
| 2023-02-10T16:33:21
| 2023-02-10T16:33:21
|
https://github.com/huggingface/datasets/issues/5304
| null |
seyong92
| false
|
[
"The [timit_asr.py](https://huggingface.co/datasets/timit_asr/blob/main/timit_asr.py) script iterates over the WAV files per split directory using this:\r\n```python\r\nwav_paths = sorted(Path(data_dir).glob(f\"**/{split}/**/*.wav\"))\r\nwav_paths = wav_paths if wav_paths else sorted(Path(data_dir).glob(f\"**/{split.upper()}/**/*.WAV\"))\r\n```\r\n\r\nCan you check that there is a directory named \"test\" somewhere in your timit data directory ?"
] |
1,464,837,251
| 5,303
|
Skip dataset verifications by default
|
closed
| 2022-11-25T18:39:09
| 2023-02-13T16:50:42
| 2023-02-13T16:43:47
|
https://github.com/huggingface/datasets/pull/5303
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5303",
"html_url": "https://github.com/huggingface/datasets/pull/5303",
"diff_url": "https://github.com/huggingface/datasets/pull/5303.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5303.patch",
"merged_at": "2023-02-13T16:43:47"
}
|
mariosasko
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"100% agree that the checksum verification is overkill and not super useful. But I think this PR would also disable the check on num_examples no ?\r\n \r\nAs a user I would like to know if the dataset I'm loading changed significantly.\r\nAnd I also think it can be useful to make sure the metadata are up to date.\r\n\r\nWhat do you think ?\r\n\r\nWe could have a default `ignore_verifications=\"ignore_checksums\"`",
"> We could have a default `ignore_verifications=\"ignore_checksums\"`\r\n\r\nAccepting multiple types (booleans and strings) at the same time is not the best design. Maybe we could define an enum for this parameter?",
"Yes an enum sounds good !",
"so we can have three verification levels, - smth like \"ignore_all\" (to skip both checksums and all other info like num_examples verification), \"ignore_checksums\" (to skip only checksums verification), and \"verify_all\" (to perform all verification)?\r\nand deprecate `ignore_verifications` param.\r\n\r\n@mariosasko if you're not going to work on this PR in the coming days, I can take over it if you want (this PR will help me with [this issue](https://github.com/huggingface/datasets/issues/5315), not super urgent though).",
"Okay, I propose deprecating `ignore_verifications` in favor of `verification_mode` (`load_dataset` already has `download_mode`; some other projects use this name for verification control). `verification_mode` would accept the following enum (or strings in the same manner as `download_mode` does):\r\n\r\n```python\r\nclass VerificationMode(enum.Enum):\r\n FULL = \"full\" # runs all verification checks \r\n BASIC = \"basic\" # default, runs only the cheap ones (skips the checksum check)\r\n NONE = \"none\" # skips all the checks\r\n```\r\n\r\nWDTY?",
"(copy paste from my message on slack)\r\n\r\nWhat do you think of a config variable in config.py to switch from one verification mode to another ? This way we don’t deprecate anything\r\n\r\nMany users are familiar with ignore_verifications=True, it might be overkill to deprecate it",
"@lhoestq So we have \"basic\" verification mode in `config.py` and continue to have `False` as a default \r\nvalue for `ignore_verifications`? That way running all verifications including checksums would not be possible without switching the config var, right? \r\n\r\nI like having a `VerificationMode` enum because it's aligned with `DownloadMode` and sounds more natural to me (`ignore_verifications` feels a bit semantically reverted but this is probably just my feeling) and it's flexible (no need to worry about `config.py`, I'm not sure that users even know it exists, wdyt?).\r\n\r\nThe usage point seems also valid to me, but cases when users are stuck with NonMatchingX errors also happen from time to time and to figure out what's wrong is non-trivial here. \r\n\r\nAs a note aside - I suggest to add instructions to the NonMatchingX error message (how to use `ignore_verifications` / `verification_mode`), this would save users who don't know about this param a lot of time.",
"Ok I see. I'm fine with the new parameter then (even though I had a small pref for the config variable) :)",
"I like the idea of an enum and the `verification_mode` parameter. \r\n\r\nIn relation with the config parameter, we could additionally add a `DEFAULT_VERIFICATION_MODE`, maybe only if users require it. Note that until now there wasn't any config parameter for a default `ignore_verifications` value: I guess people are explicitly passing `ignore_verifications=True`...\r\n\r\nAs a note aside, I like the suggestion by @polinaeterna: we could give actionable messages when verifying checksums. This could be done in other PR.",
"<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.012891 / 0.011353 (0.001538) | 0.006474 / 0.011008 (-0.004535) | 0.144038 / 0.038508 (0.105530) | 0.036151 / 0.023109 (0.013042) | 0.404366 / 0.275898 (0.128468) | 0.479988 / 0.323480 (0.156508) | 0.010219 / 0.007986 (0.002233) | 0.005319 / 0.004328 (0.000990) | 0.099705 / 0.004250 (0.095455) | 0.046639 / 0.037052 (0.009586) | 0.398997 / 0.258489 (0.140508) | 0.478431 / 0.293841 (0.184590) | 0.069125 / 0.128546 (-0.059421) | 0.019603 / 0.075646 (-0.056043) | 0.400829 / 0.419271 (-0.018443) | 0.066549 / 0.043533 (0.023016) | 0.398343 / 0.255139 (0.143204) | 0.417928 / 0.283200 (0.134728) | 0.121124 / 0.141683 (-0.020559) | 1.751513 / 1.452155 (0.299358) | 1.821239 / 1.492716 (0.328523) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.251603 / 0.018006 (0.233597) | 0.579916 / 0.000490 (0.579427) | 0.003257 / 0.000200 (0.003058) | 0.000109 / 0.000054 (0.000054) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.031502 / 0.037411 (-0.005909) | 0.134688 / 0.014526 (0.120162) | 0.152306 / 0.176557 (-0.024251) | 0.198943 / 0.737135 (-0.538192) | 0.142551 / 0.296338 (-0.153788) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.634672 / 0.215209 (0.419463) | 6.370215 / 2.077655 (4.292561) | 2.548123 / 1.504120 (1.044003) | 2.184263 / 1.541195 (0.643069) | 2.239026 / 1.468490 (0.770536) | 1.233340 / 4.584777 (-3.351437) | 5.791824 / 3.745712 (2.046112) | 5.093032 / 5.269862 (-0.176830) | 2.849833 / 4.565676 (-1.715844) | 0.143787 / 0.424275 (-0.280488) | 0.015279 / 0.007607 (0.007672) | 0.757984 / 0.226044 (0.531939) | 7.883604 / 2.268929 (5.614675) | 3.321591 / 55.444624 (-52.123033) | 2.671777 / 6.876477 (-4.204700) | 2.685215 / 2.142072 (0.543142) | 1.546709 / 4.805227 (-3.258519) | 0.247186 / 6.500664 (-6.253478) | 0.085117 / 0.075469 (0.009648) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.679809 / 1.841788 (-0.161979) | 18.528893 / 8.074308 (10.454585) | 23.168590 / 10.191392 (12.977198) | 0.277618 / 0.680424 (-0.402806) | 0.045109 / 0.534201 (-0.489092) | 0.568873 / 0.579283 (-0.010410) | 0.695017 / 0.434364 (0.260653) | 0.671024 / 0.540337 (0.130687) | 0.823817 / 1.386936 (-0.563119) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.009809 / 0.011353 (-0.001544) | 0.006890 / 0.011008 (-0.004118) | 0.099211 / 0.038508 (0.060703) | 0.035387 / 0.023109 (0.012278) | 0.507603 / 0.275898 (0.231705) | 0.535553 / 0.323480 (0.212073) | 0.007346 / 0.007986 (-0.000640) | 0.007559 / 0.004328 (0.003231) | 0.099132 / 0.004250 (0.094882) | 0.048048 / 0.037052 (0.010996) | 0.518096 / 0.258489 (0.259607) | 0.561134 / 0.293841 (0.267294) | 0.057580 / 0.128546 (-0.070966) | 0.023665 / 0.075646 (-0.051982) | 0.138409 / 0.419271 (-0.280862) | 0.061989 / 0.043533 (0.018456) | 0.510568 / 0.255139 (0.255429) | 0.552722 / 0.283200 (0.269522) | 0.115990 / 0.141683 (-0.025693) | 1.884900 / 1.452155 (0.432745) | 1.990604 / 1.492716 (0.497888) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.280638 / 0.018006 (0.262632) | 0.592837 / 0.000490 (0.592347) | 0.000465 / 0.000200 (0.000265) | 0.000078 / 0.000054 (0.000024) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.030253 / 0.037411 (-0.007158) | 0.141580 / 0.014526 (0.127054) | 0.135114 / 0.176557 (-0.041443) | 0.190003 / 0.737135 (-0.547133) | 0.160230 / 0.296338 (-0.136109) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.699762 / 0.215209 (0.484553) | 6.632344 / 2.077655 (4.554689) | 2.718803 / 1.504120 (1.214683) | 2.485294 / 1.541195 (0.944099) | 2.579889 / 1.468490 (1.111399) | 1.268795 / 4.584777 (-3.315982) | 5.777745 / 3.745712 (2.032033) | 3.232551 / 5.269862 (-2.037311) | 2.127699 / 4.565676 (-2.437977) | 0.146570 / 0.424275 (-0.277705) | 0.015971 / 0.007607 (0.008364) | 0.803181 / 0.226044 (0.577137) | 8.377192 / 2.268929 (6.108264) | 3.551242 / 55.444624 (-51.893382) | 2.865228 / 6.876477 (-4.011249) | 2.774869 / 2.142072 (0.632797) | 1.553856 / 4.805227 (-3.251371) | 0.264510 / 6.500664 (-6.236154) | 0.087918 / 0.075469 (0.012449) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.653396 / 1.841788 (-0.188391) | 18.703863 / 8.074308 (10.629555) | 22.067331 / 10.191392 (11.875939) | 0.257424 / 0.680424 (-0.422999) | 0.026448 / 0.534201 (-0.507753) | 0.550100 / 0.579283 (-0.029183) | 0.647296 / 0.434364 (0.212932) | 0.657476 / 0.540337 (0.117138) | 0.781119 / 1.386936 (-0.605817) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#8c4a9cb95f8742a2850f11d59abbef71d6c1f60c \"CML watermark\")\n",
"<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008889 / 0.011353 (-0.002464) | 0.004563 / 0.011008 (-0.006445) | 0.101627 / 0.038508 (0.063118) | 0.030526 / 0.023109 (0.007417) | 0.297175 / 0.275898 (0.021277) | 0.368454 / 0.323480 (0.044974) | 0.007246 / 0.007986 (-0.000740) | 0.003565 / 0.004328 (-0.000763) | 0.078644 / 0.004250 (0.074394) | 0.038616 / 0.037052 (0.001564) | 0.310521 / 0.258489 (0.052032) | 0.348014 / 0.293841 (0.054173) | 0.033463 / 0.128546 (-0.095083) | 0.011544 / 0.075646 (-0.064102) | 0.323281 / 0.419271 (-0.095990) | 0.040187 / 0.043533 (-0.003346) | 0.298015 / 0.255139 (0.042876) | 0.326392 / 0.283200 (0.043193) | 0.088730 / 0.141683 (-0.052952) | 1.503387 / 1.452155 (0.051233) | 1.548704 / 1.492716 (0.055988) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.185983 / 0.018006 (0.167977) | 0.451889 / 0.000490 (0.451400) | 0.001433 / 0.000200 (0.001233) | 0.000080 / 0.000054 (0.000026) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023396 / 0.037411 (-0.014015) | 0.118236 / 0.014526 (0.103710) | 0.124594 / 0.176557 (-0.051962) | 0.159089 / 0.737135 (-0.578047) | 0.129369 / 0.296338 (-0.166969) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.423161 / 0.215209 (0.207952) | 4.228211 / 2.077655 (2.150556) | 1.853862 / 1.504120 (0.349742) | 1.649471 / 1.541195 (0.108276) | 1.708631 / 1.468490 (0.240141) | 0.697456 / 4.584777 (-3.887321) | 3.473244 / 3.745712 (-0.272468) | 1.942586 / 5.269862 (-3.327275) | 1.291592 / 4.565676 (-3.274084) | 0.082758 / 0.424275 (-0.341517) | 0.012256 / 0.007607 (0.004649) | 0.528355 / 0.226044 (0.302311) | 5.277620 / 2.268929 (3.008691) | 2.299604 / 55.444624 (-53.145020) | 1.954940 / 6.876477 (-4.921537) | 2.055543 / 2.142072 (-0.086529) | 0.814723 / 4.805227 (-3.990505) | 0.149937 / 6.500664 (-6.350727) | 0.064529 / 0.075469 (-0.010941) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.266240 / 1.841788 (-0.575547) | 14.144016 / 8.074308 (6.069708) | 14.331733 / 10.191392 (4.140340) | 0.138963 / 0.680424 (-0.541461) | 0.029034 / 0.534201 (-0.505167) | 0.397325 / 0.579283 (-0.181958) | 0.405293 / 0.434364 (-0.029071) | 0.480745 / 0.540337 (-0.059592) | 0.573386 / 1.386936 (-0.813550) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.007214 / 0.011353 (-0.004139) | 0.004569 / 0.011008 (-0.006439) | 0.078718 / 0.038508 (0.040209) | 0.031104 / 0.023109 (0.007995) | 0.342562 / 0.275898 (0.066664) | 0.387802 / 0.323480 (0.064322) | 0.005378 / 0.007986 (-0.002608) | 0.003414 / 0.004328 (-0.000915) | 0.077249 / 0.004250 (0.072999) | 0.044337 / 0.037052 (0.007285) | 0.341397 / 0.258489 (0.082907) | 0.385536 / 0.293841 (0.091695) | 0.033257 / 0.128546 (-0.095289) | 0.011825 / 0.075646 (-0.063821) | 0.086723 / 0.419271 (-0.332549) | 0.045951 / 0.043533 (0.002418) | 0.340914 / 0.255139 (0.085775) | 0.367126 / 0.283200 (0.083926) | 0.096326 / 0.141683 (-0.045357) | 1.608612 / 1.452155 (0.156458) | 1.687251 / 1.492716 (0.194534) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.227595 / 0.018006 (0.209589) | 0.418502 / 0.000490 (0.418013) | 0.000392 / 0.000200 (0.000192) | 0.000059 / 0.000054 (0.000005) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.026232 / 0.037411 (-0.011179) | 0.101020 / 0.014526 (0.086494) | 0.110017 / 0.176557 (-0.066539) | 0.153497 / 0.737135 (-0.583639) | 0.110602 / 0.296338 (-0.185737) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.433789 / 0.215209 (0.218579) | 4.329350 / 2.077655 (2.251696) | 2.052136 / 1.504120 (0.548016) | 1.848457 / 1.541195 (0.307262) | 1.936791 / 1.468490 (0.468301) | 0.700609 / 4.584777 (-3.884168) | 3.391983 / 3.745712 (-0.353729) | 1.903220 / 5.269862 (-3.366642) | 1.179463 / 4.565676 (-3.386213) | 0.084025 / 0.424275 (-0.340250) | 0.012743 / 0.007607 (0.005136) | 0.536816 / 0.226044 (0.310772) | 5.420230 / 2.268929 (3.151302) | 2.507438 / 55.444624 (-52.937187) | 2.178907 / 6.876477 (-4.697570) | 2.228586 / 2.142072 (0.086514) | 0.812527 / 4.805227 (-3.992701) | 0.153382 / 6.500664 (-6.347282) | 0.069932 / 0.075469 (-0.005537) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.256861 / 1.841788 (-0.584927) | 14.309236 / 8.074308 (6.234928) | 13.740323 / 10.191392 (3.548931) | 0.142698 / 0.680424 (-0.537726) | 0.016998 / 0.534201 (-0.517203) | 0.385489 / 0.579283 (-0.193794) | 0.391515 / 0.434364 (-0.042849) | 0.472704 / 0.540337 (-0.067633) | 0.565042 / 1.386936 (-0.821894) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#4b0713ddf2e2e7129d9ccda791d265684c96675c \"CML watermark\")\n",
"This is ready for review. \r\n\r\nIf `verification_mode` is None, it defaults to `VerificationMode.BASIC` instead of `VerificationMode.NONE`, so maybe we should find a better name for the latter to avoid confusion.\r\n\r\nPS: `ignore_verifications` is still present in the `test`/`run_beam` commands for simplicity. Let me know if you think these commands should support all three modes.",
"> I would also prefer to change the name for the NONE verification mode, but don't have really good ideas in mind. maybe smth like SKIP_ALL ?\r\n\r\nI decided to go with the following names:\r\n* `no_checks` (previously `none`)\r\n* `basic_checks` (previously `basic`)\r\n* `all_checks` (previously `full`)\r\n",
"<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008900 / 0.011353 (-0.002453) | 0.004492 / 0.011008 (-0.006516) | 0.100957 / 0.038508 (0.062449) | 0.030145 / 0.023109 (0.007036) | 0.302531 / 0.275898 (0.026633) | 0.344072 / 0.323480 (0.020592) | 0.007032 / 0.007986 (-0.000953) | 0.004150 / 0.004328 (-0.000178) | 0.078272 / 0.004250 (0.074021) | 0.034142 / 0.037052 (-0.002910) | 0.310798 / 0.258489 (0.052308) | 0.350077 / 0.293841 (0.056236) | 0.034497 / 0.128546 (-0.094050) | 0.011417 / 0.075646 (-0.064230) | 0.323427 / 0.419271 (-0.095844) | 0.045664 / 0.043533 (0.002132) | 0.304688 / 0.255139 (0.049549) | 0.336591 / 0.283200 (0.053391) | 0.086116 / 0.141683 (-0.055567) | 1.519278 / 1.452155 (0.067123) | 1.576728 / 1.492716 (0.084011) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.242482 / 0.018006 (0.224476) | 0.403548 / 0.000490 (0.403058) | 0.001217 / 0.000200 (0.001017) | 0.000073 / 0.000054 (0.000018) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.023466 / 0.037411 (-0.013945) | 0.095220 / 0.014526 (0.080694) | 0.104119 / 0.176557 (-0.072438) | 0.141107 / 0.737135 (-0.596029) | 0.107236 / 0.296338 (-0.189102) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.416290 / 0.215209 (0.201081) | 4.159068 / 2.077655 (2.081413) | 1.846014 / 1.504120 (0.341894) | 1.634789 / 1.541195 (0.093594) | 1.724687 / 1.468490 (0.256196) | 0.696887 / 4.584777 (-3.887890) | 3.313861 / 3.745712 (-0.431851) | 1.907239 / 5.269862 (-3.362622) | 1.266815 / 4.565676 (-3.298861) | 0.081660 / 0.424275 (-0.342615) | 0.012290 / 0.007607 (0.004683) | 0.522866 / 0.226044 (0.296822) | 5.237356 / 2.268929 (2.968428) | 2.294645 / 55.444624 (-53.149979) | 1.946407 / 6.876477 (-4.930069) | 1.995441 / 2.142072 (-0.146632) | 0.808340 / 4.805227 (-3.996887) | 0.149670 / 6.500664 (-6.350994) | 0.065162 / 0.075469 (-0.010307) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.219476 / 1.841788 (-0.622312) | 13.868709 / 8.074308 (5.794401) | 14.115783 / 10.191392 (3.924391) | 0.149403 / 0.680424 (-0.531021) | 0.028514 / 0.534201 (-0.505686) | 0.398194 / 0.579283 (-0.181089) | 0.410898 / 0.434364 (-0.023466) | 0.485763 / 0.540337 (-0.054574) | 0.574924 / 1.386936 (-0.812012) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006906 / 0.011353 (-0.004447) | 0.004446 / 0.011008 (-0.006562) | 0.075936 / 0.038508 (0.037428) | 0.027693 / 0.023109 (0.004584) | 0.339505 / 0.275898 (0.063607) | 0.383315 / 0.323480 (0.059835) | 0.005138 / 0.007986 (-0.002847) | 0.004636 / 0.004328 (0.000308) | 0.074829 / 0.004250 (0.070578) | 0.040327 / 0.037052 (0.003274) | 0.340516 / 0.258489 (0.082027) | 0.388569 / 0.293841 (0.094729) | 0.031562 / 0.128546 (-0.096984) | 0.011585 / 0.075646 (-0.064061) | 0.084753 / 0.419271 (-0.334518) | 0.041310 / 0.043533 (-0.002223) | 0.338272 / 0.255139 (0.083133) | 0.367243 / 0.283200 (0.084043) | 0.092653 / 0.141683 (-0.049029) | 1.515973 / 1.452155 (0.063818) | 1.582869 / 1.492716 (0.090152) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.229366 / 0.018006 (0.211360) | 0.414404 / 0.000490 (0.413914) | 0.002922 / 0.000200 (0.002723) | 0.000075 / 0.000054 (0.000020) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.026391 / 0.037411 (-0.011020) | 0.106754 / 0.014526 (0.092228) | 0.110718 / 0.176557 (-0.065839) | 0.145786 / 0.737135 (-0.591350) | 0.113180 / 0.296338 (-0.183159) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.446340 / 0.215209 (0.231131) | 4.499756 / 2.077655 (2.422101) | 2.071485 / 1.504120 (0.567365) | 1.873223 / 1.541195 (0.332029) | 1.931562 / 1.468490 (0.463071) | 0.699270 / 4.584777 (-3.885507) | 3.452383 / 3.745712 (-0.293329) | 2.970630 / 5.269862 (-2.299232) | 1.300859 / 4.565676 (-3.264817) | 0.083971 / 0.424275 (-0.340304) | 0.012489 / 0.007607 (0.004882) | 0.544190 / 0.226044 (0.318146) | 5.460097 / 2.268929 (3.191169) | 2.700244 / 55.444624 (-52.744380) | 2.396694 / 6.876477 (-4.479783) | 2.376334 / 2.142072 (0.234262) | 0.812845 / 4.805227 (-3.992382) | 0.154441 / 6.500664 (-6.346223) | 0.069510 / 0.075469 (-0.005959) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.278836 / 1.841788 (-0.562952) | 14.153158 / 8.074308 (6.078850) | 13.821290 / 10.191392 (3.629898) | 0.160464 / 0.680424 (-0.519960) | 0.016742 / 0.534201 (-0.517459) | 0.379840 / 0.579283 (-0.199443) | 0.391903 / 0.434364 (-0.042461) | 0.461646 / 0.540337 (-0.078691) | 0.550691 / 1.386936 (-0.836245) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#aeb637daab938d51b8b15ad4d175d06817e99512 \"CML watermark\")\n",
"<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.009858 / 0.011353 (-0.001495) | 0.005383 / 0.011008 (-0.005625) | 0.100527 / 0.038508 (0.062019) | 0.037176 / 0.023109 (0.014067) | 0.295204 / 0.275898 (0.019306) | 0.364511 / 0.323480 (0.041031) | 0.008486 / 0.007986 (0.000500) | 0.004273 / 0.004328 (-0.000055) | 0.076538 / 0.004250 (0.072288) | 0.046250 / 0.037052 (0.009197) | 0.307102 / 0.258489 (0.048613) | 0.339313 / 0.293841 (0.045472) | 0.040783 / 0.128546 (-0.087763) | 0.012323 / 0.075646 (-0.063323) | 0.336216 / 0.419271 (-0.083055) | 0.050480 / 0.043533 (0.006947) | 0.293689 / 0.255139 (0.038550) | 0.315034 / 0.283200 (0.031834) | 0.113775 / 0.141683 (-0.027908) | 1.438738 / 1.452155 (-0.013416) | 1.499874 / 1.492716 (0.007157) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.202392 / 0.018006 (0.184386) | 0.442784 / 0.000490 (0.442295) | 0.003004 / 0.000200 (0.002804) | 0.000087 / 0.000054 (0.000033) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.027792 / 0.037411 (-0.009620) | 0.110886 / 0.014526 (0.096360) | 0.121041 / 0.176557 (-0.055515) | 0.166803 / 0.737135 (-0.570333) | 0.127617 / 0.296338 (-0.168722) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.409762 / 0.215209 (0.194553) | 4.073297 / 2.077655 (1.995643) | 1.836375 / 1.504120 (0.332255) | 1.651507 / 1.541195 (0.110312) | 1.734134 / 1.468490 (0.265644) | 0.690900 / 4.584777 (-3.893877) | 3.812045 / 3.745712 (0.066333) | 2.101378 / 5.269862 (-3.168483) | 1.438242 / 4.565676 (-3.127434) | 0.083256 / 0.424275 (-0.341020) | 0.012436 / 0.007607 (0.004829) | 0.501702 / 0.226044 (0.275658) | 5.007679 / 2.268929 (2.738751) | 2.315158 / 55.444624 (-53.129466) | 2.003934 / 6.876477 (-4.872543) | 2.154658 / 2.142072 (0.012586) | 0.831749 / 4.805227 (-3.973478) | 0.165058 / 6.500664 (-6.335606) | 0.062166 / 0.075469 (-0.013303) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.212435 / 1.841788 (-0.629353) | 15.022673 / 8.074308 (6.948365) | 14.649631 / 10.191392 (4.458239) | 0.172121 / 0.680424 (-0.508303) | 0.028791 / 0.534201 (-0.505410) | 0.440290 / 0.579283 (-0.138993) | 0.437359 / 0.434364 (0.002995) | 0.543603 / 0.540337 (0.003265) | 0.643241 / 1.386936 (-0.743695) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.007572 / 0.011353 (-0.003781) | 0.005207 / 0.011008 (-0.005801) | 0.074427 / 0.038508 (0.035919) | 0.033384 / 0.023109 (0.010275) | 0.334538 / 0.275898 (0.058640) | 0.371556 / 0.323480 (0.048076) | 0.006453 / 0.007986 (-0.001532) | 0.004010 / 0.004328 (-0.000319) | 0.073488 / 0.004250 (0.069238) | 0.048082 / 0.037052 (0.011030) | 0.337325 / 0.258489 (0.078836) | 0.395143 / 0.293841 (0.101302) | 0.036714 / 0.128546 (-0.091832) | 0.012089 / 0.075646 (-0.063557) | 0.086008 / 0.419271 (-0.333263) | 0.049277 / 0.043533 (0.005744) | 0.333848 / 0.255139 (0.078709) | 0.354003 / 0.283200 (0.070803) | 0.105012 / 0.141683 (-0.036671) | 1.450769 / 1.452155 (-0.001386) | 1.554538 / 1.492716 (0.061821) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.208407 / 0.018006 (0.190400) | 0.438778 / 0.000490 (0.438288) | 0.000399 / 0.000200 (0.000199) | 0.000059 / 0.000054 (0.000005) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.030180 / 0.037411 (-0.007232) | 0.115432 / 0.014526 (0.100906) | 0.126106 / 0.176557 (-0.050451) | 0.167508 / 0.737135 (-0.569627) | 0.130566 / 0.296338 (-0.165772) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.421408 / 0.215209 (0.206198) | 4.208492 / 2.077655 (2.130838) | 2.024177 / 1.504120 (0.520057) | 1.834356 / 1.541195 (0.293161) | 1.923234 / 1.468490 (0.454744) | 0.699548 / 4.584777 (-3.885229) | 3.933775 / 3.745712 (0.188063) | 2.124526 / 5.269862 (-3.145336) | 1.360934 / 4.565676 (-3.204742) | 0.086568 / 0.424275 (-0.337707) | 0.012351 / 0.007607 (0.004744) | 0.517431 / 0.226044 (0.291387) | 5.175428 / 2.268929 (2.906499) | 2.471031 / 55.444624 (-52.973593) | 2.131529 / 6.876477 (-4.744948) | 2.202512 / 2.142072 (0.060440) | 0.849364 / 4.805227 (-3.955863) | 0.171505 / 6.500664 (-6.329159) | 0.065864 / 0.075469 (-0.009605) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.270054 / 1.841788 (-0.571734) | 15.254502 / 8.074308 (7.180194) | 13.874969 / 10.191392 (3.683577) | 0.144131 / 0.680424 (-0.536293) | 0.017743 / 0.534201 (-0.516458) | 0.421990 / 0.579283 (-0.157293) | 0.423924 / 0.434364 (-0.010439) | 0.522560 / 0.540337 (-0.017778) | 0.626159 / 1.386936 (-0.760777) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#05bd726a575a3c1c337022424fa7d226f1a2ebee \"CML watermark\")\n",
"<details>\n<summary>Show benchmarks</summary>\n\nPyArrow==6.0.0\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.008643 / 0.011353 (-0.002710) | 0.004479 / 0.011008 (-0.006529) | 0.102372 / 0.038508 (0.063864) | 0.029703 / 0.023109 (0.006594) | 0.301479 / 0.275898 (0.025581) | 0.370970 / 0.323480 (0.047490) | 0.007044 / 0.007986 (-0.000942) | 0.004868 / 0.004328 (0.000540) | 0.079568 / 0.004250 (0.075318) | 0.035344 / 0.037052 (-0.001708) | 0.308091 / 0.258489 (0.049602) | 0.353812 / 0.293841 (0.059971) | 0.033406 / 0.128546 (-0.095140) | 0.011476 / 0.075646 (-0.064170) | 0.324343 / 0.419271 (-0.094929) | 0.040293 / 0.043533 (-0.003240) | 0.300007 / 0.255139 (0.044868) | 0.334410 / 0.283200 (0.051210) | 0.086553 / 0.141683 (-0.055130) | 1.463814 / 1.452155 (0.011659) | 1.501580 / 1.492716 (0.008864) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.198032 / 0.018006 (0.180025) | 0.409970 / 0.000490 (0.409480) | 0.001075 / 0.000200 (0.000875) | 0.000076 / 0.000054 (0.000022) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.022941 / 0.037411 (-0.014471) | 0.097320 / 0.014526 (0.082794) | 0.106445 / 0.176557 (-0.070111) | 0.139073 / 0.737135 (-0.598063) | 0.108408 / 0.296338 (-0.187930) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.419315 / 0.215209 (0.204106) | 4.199273 / 2.077655 (2.121618) | 1.877689 / 1.504120 (0.373569) | 1.670442 / 1.541195 (0.129247) | 1.735034 / 1.468490 (0.266544) | 0.694691 / 4.584777 (-3.890086) | 3.323644 / 3.745712 (-0.422069) | 2.884349 / 5.269862 (-2.385513) | 1.518882 / 4.565676 (-3.046794) | 0.082390 / 0.424275 (-0.341886) | 0.012884 / 0.007607 (0.005277) | 0.525103 / 0.226044 (0.299058) | 5.277297 / 2.268929 (3.008369) | 2.328639 / 55.444624 (-53.115985) | 1.983210 / 6.876477 (-4.893267) | 2.037985 / 2.142072 (-0.104088) | 0.809520 / 4.805227 (-3.995707) | 0.150150 / 6.500664 (-6.350514) | 0.065578 / 0.075469 (-0.009891) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.221971 / 1.841788 (-0.619817) | 13.692361 / 8.074308 (5.618052) | 13.874582 / 10.191392 (3.683190) | 0.138182 / 0.680424 (-0.542242) | 0.028618 / 0.534201 (-0.505583) | 0.395104 / 0.579283 (-0.184179) | 0.397169 / 0.434364 (-0.037195) | 0.457509 / 0.540337 (-0.082829) | 0.537275 / 1.386936 (-0.849661) |\n\n</details>\nPyArrow==latest\n\n<details>\n<summary>Show updated benchmarks!</summary>\n\n### Benchmark: benchmark_array_xd.json\n\n| metric | read_batch_formatted_as_numpy after write_array2d | read_batch_formatted_as_numpy after write_flattened_sequence | read_batch_formatted_as_numpy after write_nested_sequence | read_batch_unformated after write_array2d | read_batch_unformated after write_flattened_sequence | read_batch_unformated after write_nested_sequence | read_col_formatted_as_numpy after write_array2d | read_col_formatted_as_numpy after write_flattened_sequence | read_col_formatted_as_numpy after write_nested_sequence | read_col_unformated after write_array2d | read_col_unformated after write_flattened_sequence | read_col_unformated after write_nested_sequence | read_formatted_as_numpy after write_array2d | read_formatted_as_numpy after write_flattened_sequence | read_formatted_as_numpy after write_nested_sequence | read_unformated after write_array2d | read_unformated after write_flattened_sequence | read_unformated after write_nested_sequence | write_array2d | write_flattened_sequence | write_nested_sequence |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.006835 / 0.011353 (-0.004518) | 0.004585 / 0.011008 (-0.006423) | 0.076877 / 0.038508 (0.038369) | 0.027305 / 0.023109 (0.004196) | 0.349085 / 0.275898 (0.073187) | 0.401416 / 0.323480 (0.077936) | 0.004912 / 0.007986 (-0.003074) | 0.003315 / 0.004328 (-0.001014) | 0.075676 / 0.004250 (0.071425) | 0.038960 / 0.037052 (0.001907) | 0.346196 / 0.258489 (0.087707) | 0.403185 / 0.293841 (0.109344) | 0.032054 / 0.128546 (-0.096493) | 0.011742 / 0.075646 (-0.063905) | 0.086631 / 0.419271 (-0.332640) | 0.041633 / 0.043533 (-0.001900) | 0.343519 / 0.255139 (0.088380) | 0.385413 / 0.283200 (0.102213) | 0.091430 / 0.141683 (-0.050253) | 1.478886 / 1.452155 (0.026731) | 1.546873 / 1.492716 (0.054156) |\n\n### Benchmark: benchmark_getitem\\_100B.json\n\n| metric | get_batch_of\\_1024\\_random_rows | get_batch_of\\_1024\\_rows | get_first_row | get_last_row |\n|--------|---|---|---|---|\n| new / old (diff) | 0.167882 / 0.018006 (0.149876) | 0.396464 / 0.000490 (0.395974) | 0.003629 / 0.000200 (0.003429) | 0.000085 / 0.000054 (0.000030) |\n\n### Benchmark: benchmark_indices_mapping.json\n\n| metric | select | shard | shuffle | sort | train_test_split |\n|--------|---|---|---|---|---|\n| new / old (diff) | 0.024829 / 0.037411 (-0.012583) | 0.099607 / 0.014526 (0.085081) | 0.106187 / 0.176557 (-0.070370) | 0.142379 / 0.737135 (-0.594756) | 0.109307 / 0.296338 (-0.187032) |\n\n### Benchmark: benchmark_iterating.json\n\n| metric | read 5000 | read 50000 | read_batch 50000 10 | read_batch 50000 100 | read_batch 50000 1000 | read_formatted numpy 5000 | read_formatted pandas 5000 | read_formatted tensorflow 5000 | read_formatted torch 5000 | read_formatted_batch numpy 5000 10 | read_formatted_batch numpy 5000 1000 | shuffled read 5000 | shuffled read 50000 | shuffled read_batch 50000 10 | shuffled read_batch 50000 100 | shuffled read_batch 50000 1000 | shuffled read_formatted numpy 5000 | shuffled read_formatted_batch numpy 5000 10 | shuffled read_formatted_batch numpy 5000 1000 |\n|--------|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 0.442276 / 0.215209 (0.227067) | 4.427099 / 2.077655 (2.349444) | 2.093407 / 1.504120 (0.589287) | 1.880973 / 1.541195 (0.339778) | 1.915592 / 1.468490 (0.447102) | 0.708196 / 4.584777 (-3.876581) | 3.417649 / 3.745712 (-0.328063) | 2.859953 / 5.269862 (-2.409909) | 1.528380 / 4.565676 (-3.037297) | 0.084054 / 0.424275 (-0.340221) | 0.012585 / 0.007607 (0.004978) | 0.537614 / 0.226044 (0.311569) | 5.409915 / 2.268929 (3.140987) | 2.555853 / 55.444624 (-52.888771) | 2.195075 / 6.876477 (-4.681402) | 2.232775 / 2.142072 (0.090703) | 0.814994 / 4.805227 (-3.990233) | 0.152882 / 6.500664 (-6.347782) | 0.067467 / 0.075469 (-0.008002) |\n\n### Benchmark: benchmark_map_filter.json\n\n| metric | filter | map fast-tokenizer batched | map identity | map identity batched | map no-op batched | map no-op batched numpy | map no-op batched pandas | map no-op batched pytorch | map no-op batched tensorflow |\n|--------|---|---|---|---|---|---|---|---|---|\n| new / old (diff) | 1.306007 / 1.841788 (-0.535780) | 13.923981 / 8.074308 (5.849673) | 13.385881 / 10.191392 (3.194489) | 0.150712 / 0.680424 (-0.529712) | 0.016731 / 0.534201 (-0.517470) | 0.376557 / 0.579283 (-0.202726) | 0.379396 / 0.434364 (-0.054968) | 0.456251 / 0.540337 (-0.084087) | 0.545731 / 1.386936 (-0.841205) |\n\n</details>\n</details>\n\n![](https://cml.dev/watermark.png#cc637d107ef3e3b9948691379312a8099b6476aa \"CML watermark\")\n"
] |
1,464,778,901
| 5,302
|
Improve `use_auth_token` docstring and deprecate `use_auth_token` in `download_and_prepare`
|
closed
| 2022-11-25T17:09:21
| 2022-12-09T14:20:15
| 2022-12-09T14:17:20
|
https://github.com/huggingface/datasets/pull/5302
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5302",
"html_url": "https://github.com/huggingface/datasets/pull/5302",
"diff_url": "https://github.com/huggingface/datasets/pull/5302.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5302.patch",
"merged_at": "2022-12-09T14:17:20"
}
|
mariosasko
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,464,749,156
| 5,301
|
Return a split Dataset in load_dataset
|
closed
| 2022-11-25T16:35:54
| 2023-09-24T10:06:15
| 2023-02-21T13:13:13
|
https://github.com/huggingface/datasets/pull/5301
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5301",
"html_url": "https://github.com/huggingface/datasets/pull/5301",
"diff_url": "https://github.com/huggingface/datasets/pull/5301.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5301.patch",
"merged_at": null
}
|
lhoestq
| true
|
[
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5301). All of your documentation changes will be reflected on that endpoint.",
"Just noticed that now we have to deal with indexed & split datasets. The remaining tests are failing because one should be able to get an indexed dataset when accessing the split of a dataset made of indexed splits (right now the index is just trashed)"
] |
1,464,697,136
| 5,300
|
Use same `num_proc` for dataset download and generation
|
closed
| 2022-11-25T15:37:42
| 2022-12-07T12:55:39
| 2022-12-07T12:52:51
|
https://github.com/huggingface/datasets/pull/5300
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5300",
"html_url": "https://github.com/huggingface/datasets/pull/5300",
"diff_url": "https://github.com/huggingface/datasets/pull/5300.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5300.patch",
"merged_at": "2022-12-07T12:52:50"
}
|
mariosasko
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"I noticed this bug the other day and was going to look into it! \"Where are these processes coming from?\" ;-)"
] |
1,464,695,091
| 5,299
|
Fix xopen for Windows pathnames
|
closed
| 2022-11-25T15:35:28
| 2022-11-29T08:23:58
| 2022-11-29T08:21:24
|
https://github.com/huggingface/datasets/pull/5299
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5299",
"html_url": "https://github.com/huggingface/datasets/pull/5299",
"diff_url": "https://github.com/huggingface/datasets/pull/5299.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5299.patch",
"merged_at": "2022-11-29T08:21:24"
}
|
albertvillanova
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,464,681,871
| 5,298
|
Bug in xopen with Windows pathnames
|
closed
| 2022-11-25T15:21:32
| 2022-11-29T08:21:25
| 2022-11-29T08:21:25
|
https://github.com/huggingface/datasets/issues/5298
| null |
albertvillanova
| false
|
[] |
1,464,554,491
| 5,297
|
Fix xjoin for Windows pathnames
|
closed
| 2022-11-25T13:30:17
| 2022-11-29T08:07:39
| 2022-11-29T08:05:12
|
https://github.com/huggingface/datasets/pull/5297
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5297",
"html_url": "https://github.com/huggingface/datasets/pull/5297",
"diff_url": "https://github.com/huggingface/datasets/pull/5297.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5297.patch",
"merged_at": "2022-11-29T08:05:12"
}
|
albertvillanova
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,464,553,580
| 5,296
|
Bug in xjoin with Windows pathnames
|
closed
| 2022-11-25T13:29:33
| 2022-11-29T08:05:13
| 2022-11-29T08:05:13
|
https://github.com/huggingface/datasets/issues/5296
| null |
albertvillanova
| false
|
[] |
1,464,006,743
| 5,295
|
Extractions failed when .zip file located on read-only path (e.g., SageMaker FastFile mode)
|
closed
| 2022-11-25T03:59:43
| 2023-07-21T14:39:09
| 2023-07-21T14:39:09
|
https://github.com/huggingface/datasets/issues/5295
| null |
verdimrc
| false
|
[
"Hi ! Thanks for reporting. Indeed the lock file should be placed in a directory with write permission (e.g. in the directory where the archive is extracted).",
"I opened https://github.com/huggingface/datasets/pull/5320 to fix this - it places the lock file in the cache directory instead of trying to put in next to the ZIP where it's read-only"
] |
1,463,679,582
| 5,294
|
Support streaming datasets with pathlib.Path.with_suffix
|
closed
| 2022-11-24T18:04:38
| 2022-11-29T07:09:08
| 2022-11-29T07:06:32
|
https://github.com/huggingface/datasets/pull/5294
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5294",
"html_url": "https://github.com/huggingface/datasets/pull/5294",
"diff_url": "https://github.com/huggingface/datasets/pull/5294.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5294.patch",
"merged_at": "2022-11-29T07:06:32"
}
|
albertvillanova
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,463,669,201
| 5,293
|
Support streaming datasets with pathlib.Path.with_suffix
|
closed
| 2022-11-24T17:52:08
| 2022-11-29T07:06:33
| 2022-11-29T07:06:33
|
https://github.com/huggingface/datasets/issues/5293
| null |
albertvillanova
| false
|
[] |
1,463,053,832
| 5,292
|
Missing documentation build for versions 2.7.1 and 2.6.2
|
closed
| 2022-11-24T09:42:10
| 2022-11-24T10:10:02
| 2022-11-24T10:10:02
|
https://github.com/huggingface/datasets/issues/5292
| null |
albertvillanova
| false
|
[
"- Build docs for 2.6.2:\r\n - Commit: a6a5a1cf4cdf1e0be65168aed5a327f543001fe8\r\n - Build docs GH Action: https://github.com/huggingface/datasets/actions/runs/3539470622/jobs/5941404044\r\n- Build docs for 2.7.1:\r\n - Commit: 5ef1ab1cc06c2b7a574bf2df454cd9fcb071ccb2\r\n - Build docs GH Action: https://github.com/huggingface/datasets/actions/runs/3539574442/jobs/5941636792"
] |
1,462,983,472
| 5,291
|
[build doc] for v2.7.1 & v2.6.2
|
closed
| 2022-11-24T08:54:47
| 2022-11-24T09:14:10
| 2022-11-24T09:11:15
|
https://github.com/huggingface/datasets/pull/5291
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5291",
"html_url": "https://github.com/huggingface/datasets/pull/5291",
"diff_url": "https://github.com/huggingface/datasets/pull/5291.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5291.patch",
"merged_at": null
}
|
mishig25
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"doc versions are built https://huggingface.co/docs/datasets/index"
] |
1,462,716,766
| 5,290
|
fix error where reading breaks when batch missing an assigned column feature
|
open
| 2022-11-24T03:53:46
| 2022-11-25T03:21:54
| null |
https://github.com/huggingface/datasets/pull/5290
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5290",
"html_url": "https://github.com/huggingface/datasets/pull/5290",
"diff_url": "https://github.com/huggingface/datasets/pull/5290.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5290.patch",
"merged_at": null
}
|
eunseojo
| true
|
[
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5290). All of your documentation changes will be reflected on that endpoint."
] |
1,462,543,139
| 5,289
|
Added support for JXL images.
|
open
| 2022-11-23T23:16:33
| 2022-11-29T18:49:46
| null |
https://github.com/huggingface/datasets/pull/5289
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5289",
"html_url": "https://github.com/huggingface/datasets/pull/5289",
"diff_url": "https://github.com/huggingface/datasets/pull/5289.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5289.patch",
"merged_at": null
}
|
alexjc
| true
|
[
"I'm fine with the addition of jxl in the list of known image extensions, this way users that have the plugin can work with their JXL datasets. WDYT @mariosasko ?",
"The docs for this PR live [here](https://moon-ci-docs.huggingface.co/docs/datasets/pr_5289). All of your documentation changes will be reflected on that endpoint.",
"I think we should wait for official support from Pillow. Plus, the linked plugin doesn't support `Image.save`, which is one of the requirements for a format to be included in `IMAGE_EXTENSIONS`.\r\n\r\n@alexjc In the meantime, one option is to add these lines to the card:\r\n```python\r\nimport importlib\r\nimport datasets\r\n\r\nif \".jxl\" not in datasets.packaged_modules.imagefolder.IMAGE_EXTENSIONS:\r\n datasets.packaged_modules.imagefolder.IMAGE_EXTENSIONS.append(\".jxl\")\r\n\r\nif \"jxl\" not in datasets.packaged_modules._EXTENSION_TO_MODULE:\r\n datasets.packaged_modules._EXTENSION_TO_MODULE[\"jxl\"] = (\"imagefolder\", {})\r\n\r\nimportlib.reload(datasets.load)\r\nds = datasets.load_dataset(\"texturedesign/td01_natural-ground-textures\")\r\n```\r\nAnd you can add a note to the card that this dataset requires the \"jxlpy\" package to work. \r\n\r\nIn this case, you can also disable the viewer to avoid the discrepancy between the data displayed in the preview and the loaded data.\r\n\r\nAnother option is to define the loading script and add `jxlpy` to the list of dependencies [here](https://github.com/huggingface/datasets-server/blob/3012da62054a025467616abc14b0b46e1f11ea13/workers/first_rows/pyproject.toml#L8) to enable the viewer. This option requires more work, so let us know if you need help.",
"Thank you both for your thoughtful replies!\r\n\r\nOne questions and and update:\r\n* The jxlpy plugin does support saving, in the `_save` function of the JXLImagePlugin file. Did it not work? I'm working on the upgrade to the latest JXL, so it'd be good to know if it failed so I can fix it.\r\n* I wrote to the Pillow maintainer and the preferred solution would be to keep JXL as a separate plugin because they're a small team don't have the resources to maintain more code.\r\n\r\nWith that in mind, let me share the minimal set of features I'd need for this to work within the `datasets` library:\r\n1. Using `load_dataset()` with the HuggingFace dataset name correctly downloads the JXL files so they are available locally. Even if the `file_name` field is left intact and not loaded as a PIL image, this is the first step.\r\n2. With minimal monkey-patching, having the `load_dataset` correctly expand `file_name` into PIL `image` fields if JXL support is available.\r\n\r\nIf both of these work, then I can use HuggingFace's hub and the `datasets` library for an MVP even if not all features are there. I don't need automatic thumbnails or previews of the dataset on the server.\r\n\r\n\r\nGiven the reply from the Pillow maintainer, what solution can we come up with that works in a more permanent way than waiting for Pillow integration (which may not happen) — assuming users install the `jxlpy` plugin separately?",
"Link to my upgrade for the latest `libjxl`, pending review and merge. I tested load/save via Pillow extensively for this: https://github.com/olokelo/jxlpy/pull/13",
"After more research, here's my latest suggestion:\r\n* Depending on the build of pillow, the source (pip or conda), the platform even, certain formats may or may not be available — despite them being in the list. For example, webp support is not consistently available.\r\n* I'd suggest adding JXL to the list and simply catching the `PIL.UnidentifiedImageError` — printing a useful error message that sends them to a Wiki page to find out what to do.\r\n* On that page would be included instructions how to install support for the format and what to do for the dataset to load correctly on any platform, both with or without conda, etc.\r\n\r\nWhat do you think?",
"> The jxlpy plugin does support saving, in the _save function of the JXLImagePlugin file. Did it not work? I'm working on the upgrade to the latest JXL, so it'd be good to know if it failed so I can fix it.\r\n\r\nMy bad, I was referring to [this](https://github.com/google/brunsli/blob/2dd949e53ed05796eb44a31cc759fbf9e6c53e2f/contrib/py/jxl_library_patches/jxl_pillow.py) version of the plugin.\r\n\r\nI still think this involves too much work:\r\n* would require a new doc page\r\n* unofficial plugins have to be imported explicitly, leading to messier code on our side\r\n* etc.\r\n\r\nFor now, it seems more reasonable to create a loading script (faster than ImageFolder, as ImageFolder has to resolve the image files first) for this particular case and add `jxlpy` to the list of the `datasets-server`'s dependencies. Also, one additional advantage of this approach is that it reports if any of the modules imported in a script is missing, which is handy in your case for the plugin lib. WDYT?",
"OK, let me try it it and I'll report back.\r\n\r\nWill the JXL files (even if unknown format) be automatically downloaded if they are linked from the `.jsonl` file?\r\n\r\n(I had trouble getting that working before this patch.)",
"> Will the JXL files (even if unknown format) be automatically downloaded if they are linked from the .jsonl file?\r\n\r\nNo, they need to be downloaded explicitly.\r\n\r\nFeel free to use 🤗 Hub discussions in your dataset repo to ping us for help (our usernames are the same there)",
"Is it possible to add support for JXL files being downloaded without needing to add server-side rendering support?",
"In the loading script, data files are downloaded with `DownloadManager` (`dl_manager` in `_split_generators`), which doesn't have any requirements regarding the actual type of the downloaded files.\r\n\r\nPS: Let's use the forum or Hub discussions for further questions to avoid pinging other participants"
] |
1,462,134,067
| 5,288
|
Lossy json serialization - deserialization of dataset info
|
open
| 2022-11-23T17:20:15
| 2022-11-25T12:53:51
| null |
https://github.com/huggingface/datasets/issues/5288
| null |
anuragprat1k
| false
|
[
"Hi ! JSON is a lossy format indeed. If you want to keep the feature types or other metadata I'd encourage you to store them as well. For example you can use `dataset.info.write_to_directory` and `DatasetInfo.from_directory` to store the feature types, split info, description, license etc."
] |
1,461,971,889
| 5,287
|
Fix methods using `IterableDataset.map` that lead to `features=None`
|
closed
| 2022-11-23T15:33:25
| 2022-11-28T15:43:14
| 2022-11-28T12:53:22
|
https://github.com/huggingface/datasets/pull/5287
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5287",
"html_url": "https://github.com/huggingface/datasets/pull/5287",
"diff_url": "https://github.com/huggingface/datasets/pull/5287.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5287.patch",
"merged_at": "2022-11-28T12:53:22"
}
|
alvarobartt
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"_The documentation is not available anymore as the PR was closed or merged._",
"Maybe other options are:\r\n* Keep the `info.features` to `None` if those were initially `None`\r\n* Infer the features with pre-fetching just if the `info.features` is `None`\r\n* If the `info.features` are there, make sure that after `map` features is not `None`",
"Hi @lhoestq something that's still not clear to me is: should we infer the features always when applying a `map` if those are initially `None`, or just assume that if the features are initially `None` those should be left that way unless the user specifically sets those (or during iter)?\r\n\r\nIn this PR I'm using `from datasets.iterable_dataset import _infer_features_from_batch` to infer the features when those are `None` using pre-fetch of `self._head()`, but I'm not sure if that's the expected behavior.\r\n\r\nThanks in advance for your help!",
"Also, the PR still has some more work to do, but probably the most relevant thing to fix right now is that the `features` are being set to `None` in the functions `IterableDataset.rename_column`, `IterableDataset.rename_columns`, and `IterableDataset.remove_columns` when the `features` originally had a value. So once that's fixed maybe we can focus on improving the current `map`'s behavior, so as to avoid this from happening also when the user uses `map` directly and not through the functions mentioned above.",
"> Cool thank you ! Resolving the features can be expensive sometimes, so maybe we don't resolve the features and we can just rename/remove columns if the features are known (i.e. if they're not None). What do you think ?\r\n\r\nThanks for the feedback! Makes sense to me 👍🏻 I'll commit the comments now!",
"Already done @lhoestq, feel free to merge whenever you want! Also before merging, can you please link the following issues https://github.com/huggingface/datasets/issues/3888, https://github.com/huggingface/datasets/issues/5245, and https://github.com/huggingface/datasets/issues/5284, so that those are closed upon merge? Thanks!"
] |
1,461,908,087
| 5,286
|
FileNotFoundError: Couldn't find file at https://dumps.wikimedia.org/enwiki/20220301/dumpstatus.json
|
closed
| 2022-11-23T14:54:15
| 2024-11-23T01:16:41
| 2022-11-25T11:33:14
|
https://github.com/huggingface/datasets/issues/5286
| null |
roritol
| false
|
[
"I found a solution \r\n\r\nIf you specifically install datasets==1.18 and then run\r\n\r\nimport datasets\r\nwiki = datasets.load_dataset('wikipedia', '20200501.en')\r\nthen this should work (it worked for me.)",
"I have the same problem here but installing datasets==1.18 wont work for me\r\n",
"This works with datasets==2.14.5\r\n\r\n`>>> datasets.__version__`\r\n`'2.14.5'`\r\n`>>> ds = load_dataset(\"wikimedia/wikipedia\", \"20231101.en\")`\r\n\r\nsource:\r\nhttps://huggingface.co/datasets/wikimedia/wikipedia"
] |
1,461,521,215
| 5,285
|
Save file name in embed_storage
|
closed
| 2022-11-23T10:55:54
| 2022-11-24T14:11:41
| 2022-11-24T14:08:37
|
https://github.com/huggingface/datasets/pull/5285
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5285",
"html_url": "https://github.com/huggingface/datasets/pull/5285",
"diff_url": "https://github.com/huggingface/datasets/pull/5285.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5285.patch",
"merged_at": "2022-11-24T14:08:37"
}
|
lhoestq
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"I updated the tests, met le know if it sounds good to you now :)"
] |
1,461,519,733
| 5,284
|
Features of IterableDataset set to None by remove column
|
closed
| 2022-11-23T10:54:59
| 2025-02-07T11:36:41
| 2022-11-28T12:53:24
|
https://github.com/huggingface/datasets/issues/5284
| null |
sanchit-gandhi
| false
|
[
"Related to https://github.com/huggingface/datasets/issues/5245",
"#self-assign",
"Thanks @lhoestq and @alvarobartt!\r\n\r\nThis would be extremely helpful to have working for the Whisper fine-tuning event - we're **only** training using streaming mode, so it'll be quite important to have this feature working to make training as easy as possible!\r\n\r\n_c.f._ https://twitter.com/sanchitgandhi99/status/1592188332171493377",
"> Thanks @lhoestq and @alvarobartt!\n> \n> \n> \n> This would be extremely helpful to have working for the Whisper fine-tuning event - we're **only** training using streaming mode, so it'll be quite important to have this feature working to make training as easy as possible!\n> \n> \n> \n> _c.f._ https://twitter.com/sanchitgandhi99/status/1592188332171493377\n\nI'm almost done with at least a temporary fix to `rename_column`, `rename_columns`, and `remove_columns`, just trying to figure out how to extend it to the `map` function itself!\n\nI'll probably open the PR for review either tomorrow or Sunday hopefully! Glad I can help you and HuggingFace 🤗 ",
"Awesome - thank you so much for this PR @alvarobartt! Is much appreciated!",
"@sanchit-gandhi PR is ready and open for review at #5287, but there's still one issue I may need @lhoestq's input :hugs:",
"Let us know @sanchit-gandhi if you need a new release of `datasets` soon with this fix included :)",
"Thanks for the fix guys! We can direct people to install `datasets` from main if that's easier!",
"Hey guys, any update around this? I'm facing the same issue with a streamable dataset. ",
"Hi @asennoussi so this was already fixed and released as part of https://github.com/huggingface/datasets/releases/tag/2.8.0, so you should be able to install it as `pip install datasets==2.8.0` or just to use `pip install datasets --upgrade` to get the latest version, as of now, the https://github.com/huggingface/datasets/releases/tag/2.9.0 released last week! 🤗",
"Still facing the same issue though: \r\n```\r\nfrom datasets import IterableDatasetDict, load_dataset\r\n\r\nraw_datasets = vectorized_datasets = IterableDatasetDict()\r\n\r\n\r\nraw_datasets[\"train\"] = load_dataset(\"asennoussi/private\", split=\"train\", use_auth_token=True, streaming=True)\r\nraw_datasets[\"test\"] = load_dataset(\"asennoussi/private\", split=\"test\", use_auth_token=True, streaming=True)\r\n\r\nprint(\"Original features: \", raw_datasets['train'].features.keys())\r\n\r\n...\r\n\r\ndef prepare_dataset(batch):\r\n\r\n # load and (possibly) resample audio datato 16kHz\r\n audio = batch[\"audio\"]\r\n\r\n # compute log-Mel input features from input audio array \r\n batch[\"input_features\"] = processor.feature_extractor(audio[\"array\"], sampling_rate=audio[\"sampling_rate\"]).input_features[0]\r\n # compute input length of audio sample in seconds\r\n batch[\"input_length\"] = len(audio[\"array\"]) / audio[\"sampling_rate\"]\r\n \r\n # optional pre-processing steps\r\n transcription = batch[\"sentence\"]\r\n \r\n # encode target text to label ids\r\n batch[\"labels\"] = processor.tokenizer(transcription).input_ids\r\n batch[\"labels_length\"] = len(batch[\"labels\"])\r\n return batch\r\n...\r\nvectorized_datasets = vectorized_datasets.remove_columns(['input_length', 'labels_length']+list(next(iter(raw_datasets.values())).features))\r\nprint(\"Processed features: \", vectorized_datasets['train'].features)\r\nprint(\"First sample:\", next(iter(vectorized_datasets['train'])))\r\n\r\n```\r\n\r\nOutput: \r\n```\r\nOriginal features: dict_keys(['path', 'audio', 'sentence'])\r\nProcessed features: None\r\n```",
"Hmm weird, could you try to print\r\n\r\n```python\r\nprint(\"Processed features: \", vectorized_datasets['train'].features)\r\n```\r\n\r\nagain after iterating over the `vectorized_datasets`? In the code above, should be last line :)",
"Didn't seem to fix it: \r\n```\r\nOriginal features: dict_keys(['path', 'audio', 'sentence'])\r\nProcessed features: None\r\nProcessed features: None\r\n```",
"Actually the culprit looks to be this one: \r\n`vectorized_datasets = raw_datasets.map(prepare_dataset).with_format(\"torch\")`\r\nWhen I remove this line: `vectorized_datasets = vectorized_datasets.remove_columns(['input_length', 'labels_length']+list(next(iter(raw_datasets.values())).features))`\r\n\r\nI still get \r\n```\r\nProcessed features: None\r\n```",
"The culprit is definitely `.map` \r\nJust validated it. \r\nAny idea please? ",
"> The culprit is definitely `.map` Just validated it. Any idea please?\r\n\r\nYes, indeed `.map` losses the features, because AFAIK pre-fetching the data to infer the features is expensive and not ideal, that's part of this issue https://github.com/huggingface/datasets/issues/3888\r\n\r\nAnyway, now you can pass the `features` as a param to `.map` as follows:\r\n\r\n```python\r\nfrom datasets import Features\r\nvectorized_datasets = raw_datasets.map(\r\n prepare_dataset,\r\n features=Features(\r\n {\"path\": raw_datasets[\"train\"].info.features[\"path\"], \"audio\": raw_datasets[\"train\"].info.features[\"audio\"], \"sentence\": raw_datasets[\"train\"].info.features[\"sentence\"]}\r\n ),\r\n).with_format(\"torch\")\r\n```\r\n\r\nAlso, to let you know, when calling `.remove_columns` over an `IterableDataset`, the `features` are not lost, as well as `.rename_column` and `rename_columns` :)\r\n\r\nMore information about the latter at https://github.com/huggingface/datasets/pull/5287",
"@asennoussi alternatively you can just call `._resolve_features()` from your `IterableDataset` and it will pre-fetch the data to resolve the features, but note that feature-inference is not as accurate as if you manually specify which features and feature-types the `IterableDataset` has, as mentioned in the comment above, the alternative is to provide `features` param to `.map` :hugs:",
"Got it thanks a lot! ",
"> > The culprit is definitely `.map` Just validated it. Any idea please?\n> \n> Yes, indeed `.map` losses the features, because AFAIK pre-fetching the data to infer the features is expensive and not ideal, that's part of this issue [#3888](https://github.com/huggingface/datasets/issues/3888)\n> \n> Anyway, now you can pass the `features` as a param to `.map` as follows:\n> \n> from datasets import Features\n> vectorized_datasets = raw_datasets.map(\n> prepare_dataset,\n> features=Features(\n> {\"path\": raw_datasets[\"train\"].info.features[\"path\"], \"audio\": raw_datasets[\"train\"].info.features[\"audio\"], \"sentence\": raw_datasets[\"train\"].info.features[\"sentence\"]}\n> ),\n> ).with_format(\"torch\")\n> Also, to let you know, when calling `.remove_columns` over an `IterableDataset`, the `features` are not lost, as well as `.rename_column` and `rename_columns` :)\n> \n> More information about the latter at [#5287](https://github.com/huggingface/datasets/pull/5287)\n\nI am very late to the game, but facing the same issue still after almost 2 years. Our dataset type is IterableDatasetDict and after mapping we still get the none features. We needed the feature names to be able to in later stage use interleave_datasets() to combine two datasets together, otherwise we got errors. I just want to mention here, if somebody is using a function like prepare_dataset(batch), for whatever feature you produce here, you need to only cast these features manually to your features. Just as an example:\n\n```python\nds = IterableDatasetDict()\n\nds[\"train\"] = load_dataset(str(DATA.DIR), \"default\", split=\"train\", trust_remote_code=True, streaming=True)\n\nds = ds.cast_column(\"audio_filepath\", Audio(sampling_rate=None))\n\ndef prepare_dataset(batch):\n audio = batch[\"audio_filepath\"]\n # compute log-Mel input features from input audio array\n batch[\"input_features\"] = feature_extractor(\n audio[\"array\"], sampling_rate=audio[\"sampling_rate\"]\n ).input_features[0]\n # encode target text to label ids\n batch[\"labels\"] = tokenizer(batch[\"text\"]).input_ids\n return batch\n\nadd_new_features = Features({'input_features': Sequence(feature=Sequence(feature=Value(dtype='float32', id=None), length=-1, id=None), length=-1, id=None),\n 'labels': Sequence(feature=Value(dtype='int64', id=None), length=-1, id=None)}) # features you produce in your prepare_dataset()\n\nds = ds.map(prepare_dataset) \nds = ds.cast(add_new_features)\n```"
] |
1,460,291,003
| 5,283
|
Release: 2.6.2
|
closed
| 2022-11-22T17:36:24
| 2022-11-22T17:50:12
| 2022-11-22T17:47:02
|
https://github.com/huggingface/datasets/pull/5283
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5283",
"html_url": "https://github.com/huggingface/datasets/pull/5283",
"diff_url": "https://github.com/huggingface/datasets/pull/5283.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5283.patch",
"merged_at": "2022-11-22T17:47:02"
}
|
albertvillanova
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,460,238,928
| 5,282
|
Release: 2.7.1
|
closed
| 2022-11-22T16:58:54
| 2022-11-22T17:21:28
| 2022-11-22T17:21:27
|
https://github.com/huggingface/datasets/pull/5282
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5282",
"html_url": "https://github.com/huggingface/datasets/pull/5282",
"diff_url": "https://github.com/huggingface/datasets/pull/5282.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5282.patch",
"merged_at": "2022-11-22T17:21:27"
}
|
albertvillanova
| true
|
[] |
1,459,930,271
| 5,281
|
Support cloud storage in load_dataset
|
open
| 2022-11-22T14:00:10
| 2024-11-15T15:03:41
| null |
https://github.com/huggingface/datasets/issues/5281
| null |
lhoestq
| false
|
[
"Or for example an archive on GitHub releases! Before I added support for JXL (locally only, PR still pending) I was considering hosting my files on GitHub instead...",
"+1 to this. I would like to use 'audiofolder' with a data_dir that's on S3, for example. I don't want to upload my dataset to the Hub, but I would find all the fingerprinting/caching features useful.",
"Adding to the conversation, Dask also uses `fsspec` for this feature.\r\n\r\n[Dask: How to connect to remote data](https://docs.dask.org/en/stable/how-to/connect-to-remote-data.html)\r\n\r\nHappy to help on this feature :D ",
"+1 to this feature request since I think it also tackles my use-case. I am collaborating with a team, working with a loading script which takes some time to generate the dataset artifacts. It would be very handy to use this as a cloud cache to avoid duplicating the effort. \r\n\r\nCurrently we could use `builder.download_and_prepare(path_to_cloud_storage, storage_options, ...)` to cache the artifacts to cloud storage, but then `builder.as_dataset()` yields `NotImplementedError: Loading a dataset cached in SomeCloudFileSystem is not supported`",
"Makes sense ! If you want to load locally a dataset that you download_and_prepared on a cloud storage, you would use `load_dataset(path_to_cloud_storage)` indeed. It would download the data from the cloud storage, cache them locally, and return a `Dataset`.",
"It seems currently the `cached_path` function handles all URLs by `get_from_cache` that only supports `ftp` and `http(s)` here:\r\nhttps://github.com/huggingface/datasets/blob/b5672a956d5de864e6f5550e493527d962d6ae55/src/datasets/utils/file_utils.py#L181\r\n\r\nI guess one can add another condition that handles `s3://` or `gs://` URLs via `fsspec` here.",
"I could use this functionality, so I put together a PR using @kyamagu's suggestion to use `fsspec` in `datasets.utils.file_utils`\r\n\r\nhttps://github.com/huggingface/datasets/pull/5580",
"Thanks @dwyatte for adding support for fsspec urls\r\n\r\nLet me just reopen this since the original issue is not resolved",
"I'm not yet understanding how to use https://github.com/huggingface/datasets/pull/5580 in order to use `load_dataset(data_files=\"s3://...\")`. Any help/example would be much appreciated :) thanks! ",
"It's still not officially supported x) But you can try to update `request_etag` in `file_utils.py` to use `fsspec_head` instead of `http_head`. It is responsible of getting the ETags of the remote files for caching. This change may do the trick for S3 urls",
"Thank you for your guys help on this and merging in #5580. I manually pulled the changes to my local datasets package (datasets.utils.file_utils.py) since it only seemed to be this file that was changed in the PR and I'm getting the error: \r\nInvalidSchema: No connection adapters were found for 's3://bucket/folder/'. I'm calling load_dataset using the S3 URI. When I use the S3 URL I get HTTPError: 403 Client Error. \r\nAm I not supposed to use the S3 URI? How do I pull in the changes from this merge? I'm running datasets 2.10.1. ",
"The current implementation depends on gcsfs/s3fs being able to authenticate through some other means e.g., environmental variables. For AWS, it looks like you can set `AWS_ACCESS_KEY_ID`, `AWS_SECRET_ACCESS_KEY`, and `AWS_SESSION_TOKEN`\r\n\r\nNote that while testing this just now, I did note a discrepancy between gcsfs and s3fs that we might want to address where gcsfs passes the timeout from `storage_options` [here](https://github.com/huggingface/datasets/blob/3e6269979fc80ae8939294d26298897f0db5b84d/src/datasets/utils/file_utils.py#L333) down into the `aiohttp.ClientSession.request`, but s3fs does not handle this (tries to pass to the `aiobotocore.session.AioSession` constructor raising `TypeError: __init__() got an unexpected keyword argument 'requests_timeout'`).\r\n\r\nIt seems like some work trying to unify kwargs across different fsspec implementations, so if the plan is to pass down `storage_options`, I wonder if we should just let users control the timeout (and other kwargs) using that and if not specified, use the default?",
"> Note that while testing this just now, I did note a discrepancy between gcsfs and s3fs that we might want to address where gcsfs passes the timeout from storage_options [here](https://github.com/huggingface/datasets/blob/3e6269979fc80ae8939294d26298897f0db5b84d/src/datasets/utils/file_utils.py#L333) down into the aiohttp.ClientSession.request, but s3fs does not handle this (tries to pass to the aiobotocore.session.AioSession constructor raising TypeError: __init__() got an unexpected keyword argument 'requests_timeout').\r\n\r\n> It seems like some work trying to unify kwargs across different fsspec implementations, so if the plan is to pass down storage_options, I wonder if we should just let users control the timeout (and other kwargs) and if not specified, use the default?\r\n\r\n@lhoestq here's a small PR for this: https://github.com/huggingface/datasets/pull/5673\r\n\r\n",
"@lhoestq sorry for being a little dense here but I am very keen to use fsspec / adlfs for for a larger image dataset I have for object detection. I have to keep it on Azure storage and would also like to avoid a full download or zipping (so use `load_dataset(..., streaming=True)`. So this development is godsend :) only... I am unable to make it work.\r\n\r\nWould you expect the setup to work for:\r\n- azure blob storage\r\n- image files (not the standard formats `json`, `parquet`....\r\n\r\n? I appreciate that you mostly focus on s3 but it seems that, similar to the [remaining cloud storage functionality](https://huggingface.co/docs/datasets/filesystems#cloud-storage), it should also work for Azure blob storage.\r\n\r\n\r\nI would imagine that something like (Streaming true or false):\r\n```python\r\nd = load_dataset(\"new_dataset.py\", storage_options=storage_options, split=\"train\")\r\n```\r\nwould work with \r\n```python\r\n# new_dataset.py\r\n....\r\n_URL=\"abfs://container/image_folder``` \r\n\r\narchive_path = dl_manager.download(_URL)\r\nsplit_metadata_paths = dl_manager.download(_METADATA_URLS)\r\nreturn [\r\n datasets.SplitGenerator(\r\n name=datasets.Split.TRAIN,\r\n gen_kwargs={\r\n \"annotation_file_path\": split_metadata_paths[\"train\"],\r\n \"files\": dl_manager.iter_files(archive_path)\r\n},\r\n ),\r\n...\r\n```\r\nbut I get \r\n```\r\nTraceback (most recent call last):\r\n... ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\r\n File \"~/miniconda3/envs/hf/lib/python3.11/site-packages/datasets/load.py\", line 1797, in load_dataset\r\n builder_instance.download_and_prepare(\r\n File \"~/miniconda3/envs/hf/lib/python3.11/site-packages/datasets/builder.py\", line 890, in download_and_prepare\r\n self._download_and_prepare(\r\n File \"~/miniconda3/envs/hf/lib/python3.11/site-packages/datasets/builder.py\", line 1649, in _download_and_prepare\r\n super()._download_and_prepare(\r\n File \"~/miniconda3/envs/hf/lib/python3.11/site-packages/datasets/builder.py\", line 963, in _download_and_prepare\r\n split_generators = self._split_generators(dl_manager, **split_generators_kwargs)\r\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\r\n File \"~/.cache/huggingface/modules/datasets_modules/datasets/new_dataset/dd26a081eab90074f41fa2c821b458424fde393cc73d3d8241aca956d1fb3aa0/new_dataset_script.py\", line 56, in _split_generators\r\n archive_path = dl_manager.download(_URL)\r\n ^^^^^^^^^^^^^^^^^^^^^^^^^\r\n File \"~/miniconda3/envs/hf/lib/python3.11/site-packages/datasets/download/download_manager.py\", line 427, in download\r\n downloaded_path_or_paths = map_nested(\r\n ^^^^^^^^^^^\r\n File \"~/miniconda3/envs/hf/lib/python3.11/site-packages/datasets/utils/py_utils.py\", line 435, in map_nested\r\n return function(data_struct)\r\n ^^^^^^^^^^^^^^^^^^^^^\r\n File \"~/miniconda3/envs/hf/lib/python3.11/site-packages/datasets/download/download_manager.py\", line 453, in _download\r\n return cached_path(url_or_filename, download_config=download_config)\r\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\r\n File \"~/miniconda3/envs/hf/lib/python3.11/site-packages/datasets/utils/file_utils.py\", line 206, in cached_path\r\n raise ValueError(f\"unable to parse {url_or_filename} as a URL or as a local path\")\r\nValueError: unable to parse abfs://container/image_folder as a URL or as a local path\r\n\r\n```\r\n",
"What version of `datasets` are you using ?",
"@lhoestq \r\nhello, i still have problem with loading json from S3:\r\n\r\n\r\nstorage_options = {\r\n \"key\": xxxx,\r\n \"secret\": xxx,\r\n \"endpoint_url\": xxxx\r\n}\r\npath = 's3://xxx/xxxxxxx.json'\r\ndataset = load_dataset(\"json\", data_files=path, storage_options=storage_options)\r\n\r\nand it throws an error:\r\nTypeError: AioSession.__init__() got an unexpected keyword argument 'hf'\r\nand I use the lastest 2.14.4_dev0 version",
"Hi @lhoestq, thanks for getting back to me :) you have been busy over the summer I see... I was on `2.12.0`. I have updated to `2.14.4`. \r\n\r\nNow `d = load_dataset(\"new_dataset.py\", storage_options=storage_options, split=\"train\", streaming=True)` works for Azure blob storage (with a local data loader script) when I explicitly list all blobs (I am struggling to make `fs.ls(<path>)` work in the script to make the list available to the download manager).\r\n\r\n Any chance that it could work out-of-the-box by supplying just the image folder, not the full list of image filenames? It seems that `dl_manager.download(_URL)` always wants one or more (possibly archived) files. In my situation, where I don't want to archive or download, it would be great to just supply the folder (seems reasonably doable with fsspec).\r\n\r\nLet me know if there is anything I can do to help.\r\n\r\nThanks,",
"> Any chance that it could work out-of-the-box by supplying just the image folder, not the full list of image filenames? It seems that dl_manager.download(_URL) always wants one or more (possibly archived) files. In my situation, where I don't want to archive or download, it would be great to just supply the folder (seems reasonably doable with fsspec).\r\n\r\n@mayorblock This is not supported right now, you have to use archives or implement a way to get the list by yourself\r\n\r\n> TypeError: AioSession.init() got an unexpected keyword argument 'hf'\r\n\r\n@hjq133 Can you update `fsspec` and try again ?\r\n\r\n```python\r\npip install -U fsspec\r\n```",
"thanks for your suggestion，it works now ! ",
"I'm seeing same problem as @hjq133 with following versions:\r\n\r\n```\r\ndatasets==2.15.0\r\n(venv) ➜ finetuning-llama2 git:(main) ✗ pip freeze | grep s3fs \r\ns3fs==2023.10.0\r\n(venv) ➜ finetuning-llama2 git:(main) ✗ pip freeze | grep fsspec\r\nfsspec==2023.10.0\r\n```",
"> @lhoestq hello, i still have problem with loading json from S3:\r\n> \r\n> storage_options = { \"key\": xxxx, \"secret\": xxx, \"endpoint_url\": xxxx } path = 's3://xxx/xxxxxxx.json' dataset = load_dataset(\"json\", data_files=path, storage_options=storage_options)\r\n> \r\n> and it throws an error: TypeError: AioSession.**init**() got an unexpected keyword argument 'hf' and I use the lastest 2.14.4_dev0 version\r\n\r\nI am trying to do the same thing, but the loading is just hanging, without any error. \r\n@lhoestq is there any documentation how to load from private s3 buckets?\r\n",
"Hi ! S3 support is still experimental. It seems like there is an extra `hf` field passed to the `s3fs` storage_options that causes this error. I just check the source code of `_prepare_single_hop_path_and_storage_options` and I think you can try passing explicitly your own `storage_options={\"s3\": {...}}`. Also note that it's generally better to load datasets from HF (we run extensive tests and benchmarks for speed and robustness)",
"That worked! Thanks\r\nIt seems thought that `data_dir=...` doesn't work on s3, only `data_files`.\r\n",
"@lhoestq Would this work either with an Azure Blob Storage Container or its respective Azure Machine Learning Datastore? If yes, what would that look like in code? I've tried a couple of combinations but no success so far, on the latest version of `datasets`. I need to migrate a dataset to the Azure cloud, `load_dataset(\"path_to_data\")` worked perfectly while the files were local only. Thank you!\r\n\r\n@mayorblock would you mind sharing how you got it to work? What did you pass as `storage_options`? Would it maybe work without a custom data loader script?",
"This ticket would be of so much help.",
"@lhoestq I've been using this feature for the last year on GCS without problem, but I think we need to fix an issue with S3 and then document the supported calling patterns to reduce confusion\r\n\r\nIt looks like `datasets` uses a default `DownloadConfig` which is where some potentially unintended storage options are getting passed to fsspec\r\n\r\n```\r\nDownloadConfig(\r\n\tcache_dir=None, \r\n\tforce_download=False, \r\n\tresume_download=False, \r\n\tlocal_files_only=False, \r\n\tproxies=None, \r\n\tuser_agent=None, \r\n\textract_compressed_file=False, \r\n\tforce_extract=False, \r\n\tdelete_extracted=False, \r\n\tuse_etag=True, \r\n\tnum_proc=None, \r\n\tmax_retries=1, \r\n\ttoken=None, \r\n\tignore_url_params=False, \r\n\tstorage_options={'hf': {'token': None, 'endpoint': 'https://huggingface.co'}}, \r\n\tdownload_desc=None\r\n)\r\n```\r\n(specifically the `storage_options={'hf': {'token': None, 'endpoint': 'https://huggingface.co'}}` part)\r\n\r\n`gcsfs` is robust to the extra key in storage options for whatever reason, but `s3fs` is not (haven't dug into why). I'm unable to test `adlfs` but it looks like people here got it working\r\n\r\nIs this an issue that needs fixed with s3fs? Or can we avoid passing these default storage options in some cases?\r\n\r\nUpdate: I think probably https://github.com/huggingface/datasets/pull/6127 is where these default storage options were introduced",
"Hmm not sure, maybe it has to do with `_prepare_single_hop_path_and_storage_options` returning the \"hf\" storage options when it shouldn't",
"Also running into this issue downloading a parquet dataset from S3 (Upload worked fine using [current main branch](https://github.com/huggingface/datasets/commit/90b896193a0a315789022580eb4c80305d168d4d)).\r\n`dataset = Dataset.from_parquet('s3://path-to-file')`\r\nraises \r\n`TypeError: AioSession.__init__() got an unexpected keyword argument 'hf'`\r\nFound that the issue is introduced in [#6028](https://github.com/huggingface/datasets/commit/14f6edd9222e577dccb962ed5338b79b73502fa5#diff-8a868b8c1127c5cbe7870bbeec30bd2cbec6696a378e65b66782a6935e3f412e)\r\n\r\nWhen commenting out the __post_init__ part to set 'hf', I am able to download the dataset.",
"Can someone paste a complete example for getting this to work? Running this:\r\n\r\n```python\r\nfrom datasets import load_from_disk, load_dataset\r\nfrom os import environ\r\nfrom s3fs import S3FileSystem\r\n\r\nstorage_options = {\r\n \"key\": \"AzureDiamond\",\r\n \"secret\": \"hunter2\",\r\n \"endpoint_url\": \"https://fly.storage.tigris.dev\"\r\n}\r\n\r\nmodel_name = \"Qwen/Qwen2.5-32B\"\r\ndataset_name = \"mlabonne/FineTome-100k\"\r\n\r\ndataset = load_dataset(\r\n f\"s3://datnybakfu/model-ready/{model_name}/{dataset_name}\",\r\n storage_options=storage_options,\r\n filesystem=S3FileSystem(),\r\n streaming=True,\r\n)\r\n```\r\n\r\nI get the following error trace:\r\n\r\n<details>\r\n<summary>Stacktrace (folded)</summary\r\n\r\n```\r\n---------------------------------------------------------------------------\r\n\r\nFileNotFoundError Traceback (most recent call last)\r\n\r\n[<ipython-input-19-0ae7d5f39c46>](https://localhost:8080/#) in <cell line: 29>()\r\n 27 dataset_name = \"mlabonne/FineTome-100k\"\r\n 28 \r\n---> 29 dataset = load_dataset(\r\n 30 f\"s3://datnybakfu/model-ready/{model_name}/{dataset_name}\",\r\n 31 storage_options=storage_options,\r\n\r\n2 frames\r\n\r\n[/usr/local/lib/python3.10/dist-packages/datasets/load.py](https://localhost:8080/#) in load_dataset(path, name, data_dir, data_files, split, cache_dir, features, download_config, download_mode, verification_mode, keep_in_memory, save_infos, revision, token, streaming, num_proc, storage_options, trust_remote_code, **config_kwargs)\r\n 2130 \r\n 2131 # Create a dataset builder\r\n-> 2132 builder_instance = load_dataset_builder(\r\n 2133 path=path,\r\n 2134 name=name,\r\n\r\n[/usr/local/lib/python3.10/dist-packages/datasets/load.py](https://localhost:8080/#) in load_dataset_builder(path, name, data_dir, data_files, cache_dir, features, download_config, download_mode, revision, token, storage_options, trust_remote_code, _require_default_config_name, **config_kwargs)\r\n 1851 download_config = download_config.copy() if download_config else DownloadConfig()\r\n 1852 download_config.storage_options.update(storage_options)\r\n-> 1853 dataset_module = dataset_module_factory(\r\n 1854 path,\r\n 1855 revision=revision,\r\n\r\n[/usr/local/lib/python3.10/dist-packages/datasets/load.py](https://localhost:8080/#) in dataset_module_factory(path, revision, download_config, download_mode, dynamic_modules_path, data_dir, data_files, cache_dir, trust_remote_code, _require_default_config_name, _require_custom_configs, **download_kwargs)\r\n 1733 )\r\n 1734 else:\r\n-> 1735 raise FileNotFoundError(f\"Couldn't find any data file at {relative_to_absolute_path(path)}.\")\r\n 1736 \r\n 1737 \r\n\r\nFileNotFoundError: Couldn't find any data file at /content/s3:/datnybakfu/model-ready/Qwen/Qwen2.5-32B/mlabonne/FineTome-100k.\r\n```\r\n\r\n</details>",
"Hi, using `s3://` as first argument in load_dataset directly is not supported. You can pass a local path or a HF repository (in the form `username/dataset_name`)\r\n\r\ncc @Wauplin if you have a script to move data from S3 to HF ? I know there is this code adapted from https://github.com/fsspec/filesystem_spec/issues/909 that works well (it creates one commit per file)\r\n\r\n```python\r\nfrom multiprocessing.pool import ThreadPool\r\n\r\nimport fsspec\r\nfrom tqdm import tqdm\r\n\r\ns3_storage_options = {} # add S3 credentials here if needed, e.g. {\"key\": aws_access_key_id, \"secret\": aws_secret_access_key}\r\na = fsspec.get_mapper(\"s3://bucket/dataset_folder\", **s3_storage_options)\r\nb = fsspec.get_mapper(\"hf://datasets/username/dataset_name\")\r\n\r\ndef f(k):\r\n b[k]=a[k]\r\n\r\nwith ThreadPool(32) as p:\r\n keys = [key for key in a.keys() if key and not key.endswith(\"/\")] # ignore root and directories\r\n for _ in tqdm(p.imap_unordered(f, keys), total=len(keys)):\r\n pass\r\n```"
] |
1,459,823,179
| 5,280
|
Import error
|
closed
| 2022-11-22T12:56:43
| 2022-12-15T19:57:40
| 2022-12-15T19:57:40
|
https://github.com/huggingface/datasets/issues/5280
| null |
feketedavid1012
| false
|
[
"Hi ! Can you \r\n```python\r\nimport platform\r\nprint(platform.python_version())\r\n```\r\nto see that it returns ?",
"Hi,\n\n3.8.13\n\nGet Outlook for Android<https://aka.ms/AAb9ysg>\n________________________________\nFrom: Quentin Lhoest ***@***.***>\nSent: Tuesday, November 22, 2022 2:37:02 PM\nTo: huggingface/datasets ***@***.***>\nCc: feketedavid1012 ***@***.***>; Author ***@***.***>\nSubject: Re: [huggingface/datasets] Import error (Issue #5280)\n\n\nHi ! Can you\n\nimport platform\nprint(platform.python_version())\n\nto see that it returns ?\n\n—\nReply to this email directly, view it on GitHub<https://github.com/huggingface/datasets/issues/5280#issuecomment-1323691385>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AJW7F5YGG32W6WABYC25NJTWJTD75ANCNFSM6AAAAAASHZJ2AU>.\nYou are receiving this because you authored the thread.Message ID: ***@***.***>\n",
"Then it should work as expected if you use the same python when using `datasets`\r\n\r\nPlease make sure you're running your code in the right environment",
"It's the right environment. But in if statement I have\n\"3.8.13\" < 3.7\nAnd in the error message is Python>=3.7 which is true in my case (3.8.13 is greater then 3.7), so I don't understand my python should be below the 3.7 which case the if statement is right, but the message is wrong, or above 3.7 which case if statement is wrong, error message is right.\n\nGet Outlook for Android<https://aka.ms/AAb9ysg>\n________________________________\nFrom: Quentin Lhoest ***@***.***>\nSent: Tuesday, November 22, 2022 2:41:43 PM\nTo: huggingface/datasets ***@***.***>\nCc: feketedavid1012 ***@***.***>; Author ***@***.***>\nSubject: Re: [huggingface/datasets] Import error (Issue #5280)\n\n\nThen it should work as expected if you use the same python when using datasets\n\nPlease make sure you're running your code in the right environment\n\n—\nReply to this email directly, view it on GitHub<https://github.com/huggingface/datasets/issues/5280#issuecomment-1323697094>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AJW7F54JURTAJJWWDO2QGI3WJTERPANCNFSM6AAAAAASHZJ2AU>.\nYou are receiving this because you authored the thread.Message ID: ***@***.***>\n",
"If you're having an error then you're not running your code in the right environment."
] |
1,459,635,002
| 5,279
|
Warn about checksums
|
closed
| 2022-11-22T10:58:48
| 2022-11-23T11:43:50
| 2022-11-23T09:47:02
|
https://github.com/huggingface/datasets/pull/5279
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5279",
"html_url": "https://github.com/huggingface/datasets/pull/5279",
"diff_url": "https://github.com/huggingface/datasets/pull/5279.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5279.patch",
"merged_at": "2022-11-23T09:47:01"
}
|
lhoestq
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"I'm also in favor of disabling this by default - it's kinda impractical",
"Great, thanks for the quick turnaround on this!"
] |
1,459,574,490
| 5,278
|
load_dataset does not read jsonl metadata file properly
|
closed
| 2022-11-22T10:24:46
| 2023-02-14T14:48:16
| 2022-11-23T11:38:35
|
https://github.com/huggingface/datasets/issues/5278
| null |
065294847
| false
|
[
"Can you try to remove \"drop_labels=false\" ? It may force the loader to infer the labels instead of reading the metadata",
"Hi, thanks for responding. I tried that, but it does not change anything.",
"Can you try updating `datasets` ? Metadata support was added in `datasets` 2.4",
"Probably the issue, will report back asap!",
"Okay, now it seems to actually load the metadata and create the train_split, but it still says only returns \"image\" and \"label\", which is always 0 since all images are from same folder",
"> Can you try updating `datasets` ? Metadata support was added in `datasets` 2.4\r\n\r\nUpdate: This was the issue."
] |
1,459,388,551
| 5,277
|
Remove YAML integer keys from class_label metadata
|
closed
| 2022-11-22T08:34:07
| 2022-11-22T13:58:26
| 2022-11-22T13:55:49
|
https://github.com/huggingface/datasets/pull/5277
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5277",
"html_url": "https://github.com/huggingface/datasets/pull/5277",
"diff_url": "https://github.com/huggingface/datasets/pull/5277.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5277.patch",
"merged_at": "2022-11-22T13:55:49"
}
|
albertvillanova
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"Also note that this approach is valid when metadata keys are str, but also if they are int.\r\n- This will be helpful for any community dataset using old integer keys in their metadata",
"perfect !"
] |
1,459,363,442
| 5,276
|
Bug in downloading common_voice data and snall chunk of it to one's own hub
|
closed
| 2022-11-22T08:17:53
| 2023-07-21T14:33:10
| 2023-07-21T14:33:10
|
https://github.com/huggingface/datasets/issues/5276
| null |
capsabogdan
| false
|
[
"Sounds like one of the file is not a valid one, can you make sure you uploaded valid mp3 files ?",
"Well I just sharded the original commonVoice dataset and pushed a small chunk of it in a private rep\n\nWhat did go wrong?\n\nHolen Sie sich Outlook für iOS<https://aka.ms/o0ukef>\n________________________________\nVon: Quentin Lhoest ***@***.***>\nGesendet: Tuesday, November 22, 2022 3:03:40 PM\nAn: huggingface/datasets ***@***.***>\nCc: capsabogdan ***@***.***>; Author ***@***.***>\nBetreff: Re: [huggingface/datasets] Bug in downloading common_voice data and snall chunk of it to one's own hub (Issue #5276)\n\n\nSounds like one of the file is not a valid one, can you make sure you uploaded valid mp3 files ?\n\n—\nReply to this email directly, view it on GitHub<https://github.com/huggingface/datasets/issues/5276#issuecomment-1323727434>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ALSIFOAPAL2V4TBJTSPMAULWJTHDZANCNFSM6AAAAAASHQJ63U>.\nYou are receiving this because you authored the thread.Message ID: ***@***.***>\n",
"It should be all good then !\r\nCould you share a link to your repository for me to investigate what went wrong ?",
"https://huggingface.co/datasets/DTU54DL/common-voice-test16k\n\nAm Di., 22. Nov. 2022 um 16:43 Uhr schrieb Quentin Lhoest <\n***@***.***>:\n\n> It should be all good then !\n> Could you share a link to your repository for me to investigate what went\n> wrong ?\n>\n> —\n> Reply to this email directly, view it on GitHub\n> <https://github.com/huggingface/datasets/issues/5276#issuecomment-1323876682>,\n> or unsubscribe\n> <https://github.com/notifications/unsubscribe-auth/ALSIFOEUJRZWXAM7DYA5VJDWJTS3NANCNFSM6AAAAAASHQJ63U>\n> .\n> You are receiving this because you authored the thread.Message ID:\n> ***@***.***>\n>\n",
"I see ! This is a bug with MP3 files.\r\n\r\nWhen we store audio data in parquet, we store the bytes and the file name. From the file name extension we know if it's a WAV, an MP3 or else. But here it looks like the paths are all None.\r\n\r\nIt looks like it comes from here:\r\n\r\nhttps://github.com/huggingface/datasets/blob/7feeb5648a63b6135a8259dedc3b1e19185ee4c7/src/datasets/features/audio.py#L212\r\n\r\nCc @polinaeterna maybe we should simply put the file name instead of None values ?",
"@lhoestq I remember we wanted to avoid storing redundant data but maybe it's not that crucial indeed to store one more string value. \r\nOr we can store paths only for mp3s, considering that for other formats we don't have such a problem with reading from bytes without format specified. ",
"It doesn't cost much to always store the file name IMO",
"thanks for the help!\n\ncan I do anything on my side? we are doing a DL project and we need the\ndata really quick.\n\nthanks\nbogdan\n\n> Message ID: ***@***.***>\n>\n",
"I opened a pull requests here: https://github.com/huggingface/datasets/pull/5285, we'll do a new release soon with this fix.\r\n\r\nOtherwise if you're really in a hurry you can install `datasets` from this PR",
"[image: image.png]\n\n> Message ID: ***@***.***>\n>\n",
"any idea on what's going wrong here?\n\nthanks\n\nAm So., 27. Nov. 2022 um 13:53 Uhr schrieb Bogdan Capsa <\n***@***.***>:\n\n> [image: image.png]\n>\n>> Message ID: ***@***.***>\n>>\n>\n",
"hi @capsabogdan! \r\ncould you please share more specifically what problem do you have now?",
"I have attached this screenshot above . can u pls help? So can not pip from pull request\r\n\r\n\r\n",
"The pull request has been merged on `main`.\r\nYou can install `datasets` from `main` using\r\n```\r\npip install git+https://github.com/huggingface/datasets.git\r\n```",
"I've tried to load this dataset DTU54DL/common-voice-test16k, but am\ngetting the same error.\n\nSo the bug fix will fix only if I upload a new dataset, or also loading\npreviously uploaded datasets?\n\nthanks\n\nAm Mo., 28. Nov. 2022 um 19:51 Uhr schrieb Quentin Lhoest <\n***@***.***>:\n\n> The pull request has been merged on main.\n> You can install datasets from main using\n>\n> pip install git+https://github.com/huggingface/datasets.git\n>\n> —\n> Reply to this email directly, view it on GitHub\n> <https://github.com/huggingface/datasets/issues/5276#issuecomment-1329587334>,\n> or unsubscribe\n> <https://github.com/notifications/unsubscribe-auth/ALSIFOCNYYIGHM2EX3ZIO6DWKT5MXANCNFSM6AAAAAASHQJ63U>\n> .\n> You are receiving this because you were mentioned.Message ID:\n> ***@***.***>\n>\n",
"> So the bug fix will fix only if I upload a new dataset, or also loading\r\npreviously uploaded datasets?\r\n\r\nYou have to reupload the dataset, sorry for the inconvenience",
"thank you so much for the help! works like a charm!\n\nAm Di., 29. Nov. 2022 um 12:15 Uhr schrieb Quentin Lhoest <\n***@***.***>:\n\n> So the bug fix will fix only if I upload a new dataset, or also loading\n> previously uploaded datasets?\n>\n> You have to reupload the dataset, sorry for the inconvenience\n>\n> —\n> Reply to this email directly, view it on GitHub\n> <https://github.com/huggingface/datasets/issues/5276#issuecomment-1330468393>,\n> or unsubscribe\n> <https://github.com/notifications/unsubscribe-auth/ALSIFOBKEFZO57BAKY4IGW3WKXQUZANCNFSM6AAAAAASHQJ63U>\n> .\n> You are receiving this because you were mentioned.Message ID:\n> ***@***.***>\n>\n"
] |
1,459,358,919
| 5,275
|
YAML integer keys are not preserved Hub server-side
|
closed
| 2022-11-22T08:14:47
| 2023-01-26T10:52:35
| 2023-01-26T10:40:21
|
https://github.com/huggingface/datasets/issues/5275
| null |
albertvillanova
| false
|
[
"@huggingface/datasets if you agree, I can make the bulk edit on the Hub to fix integer keys into strings.",
"Ok for me, and we can merge (internal) https://github.com/huggingface/moon-landing/pull/4609",
"FYI there are still 2k+ weekly users on `datasets` 2.6.1 which doesn't support the string label format for class labels. And among those, some are using datasets with class labels like imdb (60 users), conllpp (40), msra_ner (40), peoples_daily_enr (40), weibo_ner (30), conll2003 (20), etc. And renaming to string would break these users code.",
"but isn't `datasets 2.6.1` downloading files from the Hub with the corresponding tag? I thought we had something like this before",
"We're using `main` as models do. Some datasets need to be updated from time to time, e.g. when a link to download the data is dead.\r\n\r\nBut yea a year ago we had those tags, we just ended up not using them",
"I opened https://github.com/huggingface/datasets/issues/5406 to communicate on this. Let me know what you think, and if it sounds good to you I can pin this issue",
"So, is it OK to make the bulk edit on the Hub now or should we wait longer? If the latter, how long?",
"I think we can do it. If you want to be extra cautious you can do it for all datasets except imdb and conllpp for now which are actively used by 2.6.1 users. For those two we can keep the YAML like this for some more time, or alternatively use the old dataset_infos.json file",
"The bulk edit of canonical datasets (except imdb and conllpp) is running. \r\n\r\nSee e.g.: https://huggingface.co/datasets/acronym_identification/discussions/3\r\n\r\nEDITED: \r\nDone, except for \"universal_morphologies\", where I get\r\n```\r\nHTTPError: 413 Client Error: Payload Too Large for url: https://huggingface.co/api/validate-yaml\r\n```\r\n\r\nAlso not done for the datasets missing matadata \"dataset_info\":\r\n- mc4: https://huggingface.co/datasets/mc4/discussions/3\r\n- the_pile: https://huggingface.co/datasets/the_pile/discussions/6\r\n- timit_asr: https://huggingface.co/datasets/timit_asr/discussions/1",
"Thank you !",
"@lhoestq, there are 6 community datasets with YAML integer keys in their `dataset_info` `class_label`:\r\n- indonlp/indonlu\r\n- rcds/swiss_judgment_prediction\r\n- Jean-Baptiste/wikiner_fr\r\n- Bingsu/Cat_and_Dog\r\n- taskydata/tasky_or_not\r\n- RCC-MSU/collection3\r\n\r\nMaybe we could open a PR on them as well?",
"Let's do this then:\r\n\r\n- [x] [indonlp/indonlu](https://huggingface.co/datasets/indonlp/indonlu/discussions/3)\r\n- [x] rcds/swiss_judgment_prediction\r\n- [x] Jean-Baptiste/wikiner_fr\r\n- [x] Bingsu/Cat_and_Dog -> merged\r\n- [x] taskydata/tasky_or_not (was already using quotes)\r\n- [x] RCC-MSU/collection3\r\n\r\nEDIT: all done :)",
"@lhoestq I was not asking you to do it, but asking if you agree me to do it... :man_facepalming: \r\nAs I self-assigned this issue... :sweat_smile: "
] |
1,458,646,455
| 5,274
|
load_dataset possibly broken for gated datasets?
|
closed
| 2022-11-21T21:59:53
| 2023-05-27T00:06:14
| 2022-11-28T02:50:42
|
https://github.com/huggingface/datasets/issues/5274
| null |
TristanThrush
| false
|
[
"@BradleyHsu",
"Btw, thanks very much for finding the hub rollback temporary fix and bringing the issue to our attention @KhoomeiK!",
"I see the same issue when calling `load_dataset('poloclub/diffusiondb', 'large_random_1k')` with `datasets==2.7.1` and `huggingface-hub=0.11.0`. No issue with `datasets=2.6.1` and `huggingface_hub==0.10.1`.\r\n\r\nhttps://github.com/poloclub/diffusiondb/issues/7",
"I fixed my issue by specifying `repo_type` in `hf_hub_url()`. https://github.com/poloclub/diffusiondb/commit/9eb91c79aaca98b0515a0ce45778b8af65b84652\r\n\r\nI opened a PR on the Winoground's repo: https://huggingface.co/datasets/facebook/winoground/discussions/2",
"This is a bug in the script, indeed. The most robust fix is to use a relative path instead of `hf_hub_url`, which does not depend on `huggingface_hub`'s version 🙂. I've opened a PR here: https://huggingface.co/datasets/facebook/winoground/discussions/3.",
"Awesome, big thanks to both @xiaohk and @mariosasko!",
"so, if i reproduce the bug, what should i do ? with huggingface_hub0.13.3 dataset2.6.1",
"huggingface_hub.utils._validators.HFValidationError: Repo id must be in the form 'repo_name' or 'namespace/repo_name':\r\n\r\n tokenizer = AutoTokenizer.from_pretrained(ARGS.model_path, trust_remote_code=True)\r\n\r\nPlease handle automatically for local path and repo name inside, otherwise users always get confused about this",
"I think I'm also hitting this error, trying to load a model from a local path."
] |
1,458,018,050
| 5,273
|
download_mode="force_redownload" does not refresh cached dataset
|
open
| 2022-11-21T14:12:43
| 2022-11-21T14:13:03
| null |
https://github.com/huggingface/datasets/issues/5273
| null |
nomisto
| false
|
[] |
1,456,940,021
| 5,272
|
Use pyarrow Tensor dtype
|
open
| 2022-11-20T15:18:41
| 2024-11-11T03:03:17
| null |
https://github.com/huggingface/datasets/issues/5272
| null |
franz101
| false
|
[
"Hi ! We're using the Arrow format for the datasets, and PyArrow tensors are not part of the Arrow format AFAIK:\r\n\r\n> There is no direct support in the arrow columnar format to store Tensors as column values.\r\n\r\nsource: https://github.com/apache/arrow/issues/4802#issuecomment-508494694",
"@wesm @rok its been around three years. any updates, regarding dataset arrow tensor support? 🙏 I know you must be very busy, would appreciate to learn what is the state of art. I saw the PR is still open [#8510](https://github.com/apache/arrow/pull/8510)",
"Hey @franz101 & @lhoestq!\r\nThere is a plan and a PR to create an [ExtensionArray of Tensors](https://github.com/apache/arrow/pull/8510) of equal sizes as well as a plan to do the same for Tensors of different sizes [ARROW-8714](https://issues.apache.org/jira/browse/ARROW-8714).",
"The work stalled a little because it was not clear where TensorArray would live. However Arrow community recently agreed to make a [well-known-extension-type document](https://lists.apache.org/thread/sxd5fhc42hb6svs79t3fd79gkqj83pfh) and I would like https://github.com/apache/arrow/pull/8510 to land there and add an implementation to C++/Python + another language. Is that something you would find beneficial to you?",
"that is a great update, thank you.\r\nit looks like this feature would benefit datasets implementation of [ArrayExtensionArray](https://github.com/huggingface/datasets/blob/9f2ff14673cac1f1ad56d80221a793f5938b68c7/src/datasets/features/features.py#L585-L641). Is that correct @eladsegal @lhoestq?\r\n\r\n",
"TensorArray sounds great ! Looking forward to it :)\r\n\r\nWe've had our own ExtensionArray for fixed shape tensors for a while now, hoping to see something more standardized by the arrow community.\r\n\r\nAlso super interested in the extension array for tensors of different sizes cc @mariosasko ",
"[FixedShapeTensor ExtensionType](https://github.com/apache/arrow/pull/8510) was merged and will be in Arrow 12.0.0 (release is planned mid April).\r\n",
"@rok Thanks for keeping us updated! I think it's best to introduce a new feature type that would use this extension type under the hood. I'll create an issue to discuss the design with the community in the coming days.\r\n\r\nAlso, is there a tentative time frame for the variable-shape Tensor extension type?",
"@mariosasko please tag me in the discussion, perhaps I can contribute.\r\n\r\nAs for the [variable shape tensor array](https://github.com/apache/arrow/issues/24868) - I'd be interested in working on it but didn't see much interest in community yet. Are you saying `huggingface/datasets` could use it?",
"pyarrow 12 is out 🎉, will have a look if I can work on it for the ExtensionArray",
"I think these two issues need to be fixed first on the Arrow side before adding the tensor feature type here: https://github.com/apache/arrow/issues/35573 and https://github.com/apache/arrow/issues/35599.\r\n\r\n@rok We've had a couple of requests for supporting variable-shape tensors on the forum/GH, but I did not manage to find the concrete issues using the search. TF/TFDS (and PyTorch with the `nested_tensor` API) support them, so it makes sense for us to do the same eventually (the Ray project has an [extension](https://github.com/ray-project/ray/blob/42a8d1489b37243f203120899a23d919dc85bf2a/python/ray/air/util/tensor_extensions/arrow.py#L634) type to support this case)",
"> @rok We've had a couple of requests for supporting variable-shape tensors on the forum/GH, but I did not manage to find the concrete issues using the search. TF/TFDS (and PyTorch with the `nested_tensor` API) support them, so it makes sense for us to do the same eventually (the Ray project has an [extension](https://github.com/ray-project/ray/blob/42a8d1489b37243f203120899a23d919dc85bf2a/python/ray/air/util/tensor_extensions/arrow.py#L634) type to support this case)\r\n\r\nThat does make sense indeed. We should probably also be careful about memory layout to enable zero-copy interface to TF/PyTorch.",
"So there is no way we can use [pyarrow.Tensor](https://arrow.apache.org/docs/python/generated/pyarrow.Tensor.html#pyarrow.Tensor) ?",
"Not with with the Arrow format, and therefore not in `datasets`. But they released a new [FixedShapeTensorArray](https://arrow.apache.org/docs/python/extending_types.html#fixed-size-tensor) to store tensors in Arrow format. We plan to support this in `datasets` at one point !",
"There is also an open issue to enable the conversion of `pyarrow.Tensor` to `pyarrow.FixedShapeTensorType`: https://github.com/apache/arrow/issues/35068. This way one could indirectly use `pyarrow.Tensor` in Arrow format.",
"We started a [mailing list discussion](https://lists.apache.org/thread/qc9qho0fg5ph1dns4hjq56hp4tj7rk1k) about potential `VariableShapeTensor` extension array, please check it out and give feedback. For more details here's also a PR https://github.com/apache/arrow/pull/37166.",
"Kindly ask what's the recent progress?"
] |
1,456,807,738
| 5,271
|
Fix #5269
|
closed
| 2022-11-20T07:50:49
| 2022-11-21T15:07:19
| 2022-11-21T15:06:38
|
https://github.com/huggingface/datasets/pull/5271
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5271",
"html_url": "https://github.com/huggingface/datasets/pull/5271",
"diff_url": "https://github.com/huggingface/datasets/pull/5271.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5271.patch",
"merged_at": null
}
|
Freed-Wu
| true
|
[
"See <https://github.com/huggingface/datasets/issues/5269>"
] |
1,456,508,990
| 5,270
|
When len(_URLS) > 16, download will hang
|
open
| 2022-11-19T14:27:41
| 2022-11-21T15:27:16
| null |
https://github.com/huggingface/datasets/issues/5270
| null |
Freed-Wu
| false
|
[
"It can fix the bug temporarily.\r\n```python\r\nfrom datasets import DownloadConfig\r\nconfig = DownloadConfig(num_proc=8)\r\nIn [5]: dataset = load_dataset('Freed-Wu/kodak', split='test', download_config=config)\r\nDownloading and preparing dataset kodak/default to /home/wzy/.cache/huggingface/datasets/Freed-Wu___kodak/default/0.0.1/6cf51f2b3d686d24a33fe86945f9e16802def212325f9345cf3cbb1b9f5f4a57...\r\nDownloading data files #4: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:02<00:00, 1.39obj/s]\r\nDownloading data files #2: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:02<00:00, 1.38obj/s]\r\nDownloading data files #3: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:02<00:00, 1.13obj/s]\r\nDownloading data files #7: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:02<00:00, 1.09obj/s]\r\nDownloading data files #5: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:02<00:00, 1.08obj/s]\r\nDownloading data files #0: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:02<00:00, 1.08obj/s]\r\nDownloading data files #1: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:10<00:00, 3.36s/obj]\r\nDownloading data: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 492k/492k [00:01<00:00, 253kB/s]\r\nDownloading data files #6: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:13<00:00, 4.63s/obj]\r\nExtracting data files #0: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:00<00:00, 1407.17obj/s]\r\nExtracting data files #1: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:00<00:00, 1325.91obj/s]\r\nExtracting data files #3: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:00<00:00, 1524.46obj/s]\r\nExtracting data files #2: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:00<00:00, 1404.66obj/s]\r\nExtracting data files #4: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:00<00:00, 1538.63obj/s]\r\nExtracting data files #6: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:00<00:00, 1711.73obj/s]\r\nExtracting data files #7: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:00<00:00, 2144.33obj/s]\r\nExtracting data files #5: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 3/3 [00:00<00:00, 1964.85obj/s]\r\nDataset kodak downloaded and prepared to /home/wzy/.cache/huggingface/datasets/Freed-Wu___kodak/default/0.0.1/6cf51f2b3d686d24a33fe86945f9e16802def212325f9345cf3cbb1b9f5f4a57. Subsequent calls will reuse this data.\r\n```",
"Thanks for reporting ! This sounds like an issue with python multiprocessing. If we switch to multithreading for the downloads it should be much more robust - let me know if this is something you'd like to contribute, I'd be happy to help and give you some pointers",
"> an issue with python multiprocessing\r\n\r\nIf it is an issue with multiprocessing, should we report it to upstream?",
"Debugging this would require quite some work in my opinion, and I've often failed to make reproducible examples, since it's pretty correlated to one's environment + hardware. So I wouldn't spend too much time on this unless we manage to reproduce this on another machine consistently.\r\n\r\nInstead I'd encourage a more pragmatic fix that is: not create tons of processes (on regular machines it may slow things down anyway), and instead use multithreading by default.",
"I am not expert of python. I hear about python has GIL, which result in multi processing is worse than multi threading. So I am not sure if this change makes sense?\r\n\r\nAnd if this is a bug of multi processing, why not report to upstream and let them fix? And even if change it to multi threading, how can we make sure it can truly fix this problem?",
"Just my 2c. No offense.",
"> Just my 2c. No offense.\r\n\r\nsure np ^^\r\n\r\n> I hear about python has GIL, which result in multi processing is worse than multi threading. So I am not sure if this change makes sense?\r\n\r\nHere the bottleneck speed is the bandwidth used to download the files. When downloading, the GIL is released, so multithreading gives the same speed as multiprocessing.\r\n\r\n> And if this is a bug of multi processing, why not report to upstream and let them fix?\r\n\r\nUsually to fix a bug it's important to be able to reproduce it. This way you can share it, experiment with it, and then make sure it's fixed. Here I'm afraid it's not easy to reproduce. Though I think that spawning too many processes for your machine can lead to this kind of issues.\r\n\r\n> And even if change it to multi threading, how can we make sure it can truly fix this problem?\r\n\r\nMultithreading is more robust in python because IIRC there are less locks involved which are often the cause of code hanging for no reason."
] |
1,456,485,799
| 5,269
|
Shell completions
|
closed
| 2022-11-19T13:48:59
| 2022-11-21T15:06:15
| 2022-11-21T15:06:14
|
https://github.com/huggingface/datasets/issues/5269
| null |
Freed-Wu
| false
|
[
"I don't think we need completion on the datasets-cli, since we're mainly developing huggingface-cli",
"I see."
] |
1,455,633,978
| 5,268
|
Sharded save_to_disk + multiprocessing
|
closed
| 2022-11-18T18:50:01
| 2022-12-14T18:25:52
| 2022-12-14T18:22:58
|
https://github.com/huggingface/datasets/pull/5268
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5268",
"html_url": "https://github.com/huggingface/datasets/pull/5268",
"diff_url": "https://github.com/huggingface/datasets/pull/5268.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5268.patch",
"merged_at": "2022-12-14T18:22:58"
}
|
lhoestq
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._",
"Added both num_shards and max_shard_size in push_to_hub/save_to_disk. Will take care of updating the tests later",
"It's ready for a final review @mariosasko and @albertvillanova, let me know what you think :)",
"Took your comments into account, and also changed `iflatmap_unordered` to take an iterable of kwargs to make the code more redable :)"
] |
1,455,466,464
| 5,267
|
Fix `max_shard_size` docs
|
closed
| 2022-11-18T16:55:22
| 2022-11-18T17:28:58
| 2022-11-18T17:25:27
|
https://github.com/huggingface/datasets/pull/5267
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5267",
"html_url": "https://github.com/huggingface/datasets/pull/5267",
"diff_url": "https://github.com/huggingface/datasets/pull/5267.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5267.patch",
"merged_at": "2022-11-18T17:25:26"
}
|
lhoestq
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,455,281,310
| 5,266
|
Specify arguments as keywords in librosa.reshape to avoid future errors
|
closed
| 2022-11-18T14:58:47
| 2022-11-21T15:45:02
| 2022-11-21T15:41:57
|
https://github.com/huggingface/datasets/pull/5266
|
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/5266",
"html_url": "https://github.com/huggingface/datasets/pull/5266",
"diff_url": "https://github.com/huggingface/datasets/pull/5266.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/5266.patch",
"merged_at": "2022-11-21T15:41:57"
}
|
polinaeterna
| true
|
[
"_The documentation is not available anymore as the PR was closed or merged._"
] |
1,455,274,864
| 5,265
|
Get an IterableDataset from a map-style Dataset
|
closed
| 2022-11-18T14:54:40
| 2023-02-01T16:36:03
| 2023-02-01T16:36:03
|
https://github.com/huggingface/datasets/issues/5265
| null |
lhoestq
| false
|
[
"I think `stream` could be misleading since the data is not being streamed from remote endpoints (one could think that's the case when they see `load_dataset` followed by `stream`). Hence, I prefer the second option.\r\n\r\nPS: When we resolve https://github.com/huggingface/datasets/issues/4542, we could add `as_tf_dataset` to the API for consistency and deprecate `to_tf_dataset`."
] |
1,455,252,906
| 5,264
|
`datasets` can't read a Parquet file in Python 3.9.13
|
closed
| 2022-11-18T14:44:01
| 2023-05-07T09:52:59
| 2022-11-22T11:18:08
|
https://github.com/huggingface/datasets/issues/5264
| null |
loubnabnl
| false
|
[
"Could you share the full stack trace please ?\r\n\r\n\r\nCan you also try running this code ? It can be useful to determine if the issue comes from `datasets` or `fsspec` (streaming) or `pyarrow` (parquet reading):\r\n```python\r\nds = load_dataset(\"parquet\", data_files=a_parquet_file_url, use_auth_token=True)\r\n```",
"Here's the full trace\r\n```\r\nTraceback (most recent call last):\r\n File \"/home/loubna_huggingface_co/load.py\", line 15, in <module>\r\n ds_all = load_dataset(\"bigcode/the-stack-dedup-pjj\", data_dir=\"data/java\",use_auth_token=True, split=\"train\", revision=\"v1.1.a1\")\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/load.py\", line 1742, in load_dataset\r\n builder_instance.download_and_prepare(\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/builder.py\", line 814, in download_and_prepare\r\n self._download_and_prepare(\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/builder.py\", line 905, in _download_and_prepare\r\n self._prepare_split(split_generator, **prepare_split_kwargs)\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/builder.py\", line 1502, in _prepare_split\r\n for key, table in logging.tqdm(\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/tqdm/std.py\", line 1195, in __iter__\r\n for obj in iterable:\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/packaged_modules/parquet/parquet.py\", line 67, in _generate_tables\r\n parquet_file = pq.ParquetFile(f)\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/pyarrow/parquet/__init__.py\", line 286, in __init__\r\n self.reader.open(\r\n File \"pyarrow/_parquet.pyx\", line 1227, in pyarrow._parquet.ParquetReader.open\r\n File \"pyarrow/error.pxi\", line 100, in pyarrow.lib.check_status\r\npyarrow.lib.ArrowInvalid: Parquet magic bytes not found in footer. Either the file is corrupted or this is not a parquet file.\r\n```\r\n\r\nwhen running\r\n```python\r\nds = load_dataset(\"parquet\", data_files=\"https://huggingface.co/datasets/bigcode/the-stack-dedup-pjj/blob/v1.1.a1/data/java/data_0000.parquet\", use_auth_token=True)\r\n```\r\nI get 401 error, but that's the case for the python subset too which I can load properly\r\n```\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/load.py\", line 1719, in load_dataset\r\n builder_instance = load_dataset_builder(\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/load.py\", line 1497, in load_dataset_builder\r\n dataset_module = dataset_module_factory(\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/load.py\", line 1134, in dataset_module_factory\r\n return PackagedDatasetModuleFactory(\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/load.py\", line 707, in get_module\r\n data_files = DataFilesDict.from_local_or_remote(\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/data_files.py\", line 795, in from_local_or_remote\r\n DataFilesList.from_local_or_remote(\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/data_files.py\", line 764, in from_local_or_remote\r\n origin_metadata = _get_origin_metadata_locally_or_by_urls(data_files, use_auth_token=use_auth_token)\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/data_files.py\", line 710, in _get_origin_metadata_locally_or_by_urls\r\n return thread_map(\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/tqdm/contrib/concurrent.py\", line 94, in thread_map\r\n return _executor_map(ThreadPoolExecutor, fn, *iterables, **tqdm_kwargs)\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/tqdm/contrib/concurrent.py\", line 76, in _executor_map\r\n return list(tqdm_class(ex.map(fn, *iterables, **map_args), **kwargs))\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/tqdm/std.py\", line 1183, in __iter__\r\n for obj in iterable:\r\n File \"/opt/conda/envs/venv/lib/python3.9/concurrent/futures/_base.py\", line 609, in result_iterator\r\n yield fs.pop().result()\r\n File \"/opt/conda/envs/venv/lib/python3.9/concurrent/futures/_base.py\", line 446, in result\r\n return self.__get_result()\r\n File \"/opt/conda/envs/venv/lib/python3.9/concurrent/futures/_base.py\", line 391, in __get_result\r\n raise self._exception\r\n File \"/opt/conda/envs/venv/lib/python3.9/concurrent/futures/thread.py\", line 58, in run\r\n result = self.fn(*self.args, **self.kwargs)\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/data_files.py\", line 701, in _get_single_origin_metadata_locally_or_by_urls\r\n return (request_etag(data_file, use_auth_token=use_auth_token),)\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/utils/file_utils.py\", line 411, in request_etag\r\n response.raise_for_status()\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/requests/models.py\", line 960, in raise_for_status\r\n raise HTTPError(http_error_msg, response=self)\r\nrequests.exceptions.HTTPError: 401 Client Error: Unauthorized for url: https://huggingface.co/datasets/bigcode/the-stack-dedup-pjj/blob/v1.1.a1/data/python/data_0000.parquet```",
"Can you check you used the right token ? You shouldn't get a 401 using your token",
"I checked it’s the right token, when loading the full dataset I get the error after data extraction so I can access the files. \r\n```\r\nDownloading and preparing dataset parquet/bigcode--the-stack-dedup-pjj to /home/loubna_huggingface_co/.cache/huggingface/datasets/bigcode___parquet/bigcode--the-stack-dedup-pjj-872ffac7f4bb46ca/0.0.0/2a3b91fbd88a2c90d1dbbb32b460cf621d31bd5b05b934492fdef7d8d6f236ec...\r\nDownloading data files: 100%|██████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:00<00:00, 22.38it/s]\r\nExtracting data files: 100%|███████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:00<00:00, 49.91it/s]\r\nTraceback (most recent call last):\r\n File \"/home/loubna_huggingface_co/load_ds.py\", line 5, in <module>\r\n ds = load_dataset(\"bigcode/the-stack-dedup-pjj\", data_dir=\"data/java\", use_auth_token=True,split=\"train\", revision=\"v1.1.a1\")\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/load.py\", line 1742, in load_dataset\r\n builder_instance.download_and_prepare(\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/builder.py\", line 814, in download_and_prepare\r\n self._download_and_prepare(\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/builder.py\", line 905, in _download_and_prepare\r\n self._prepare_split(split_generator, **prepare_split_kwargs)\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/builder.py\", line 1502, in _prepare_split\r\n for key, table in logging.tqdm(\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/tqdm/std.py\", line 1195, in __iter__\r\n for obj in iterable:\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/datasets/packaged_modules/parquet/parquet.py\", line 67, in _generate_tables\r\n parquet_file = pq.ParquetFile(f)\r\n File \"/opt/conda/envs/venv/lib/python3.9/site-packages/pyarrow/parquet/__init__.py\", line 286, in __init__\r\n self.reader.open(\r\n File \"pyarrow/_parquet.pyx\", line 1227, in pyarrow._parquet.ParquetReader.open\r\n File \"pyarrow/error.pxi\", line 100, in pyarrow.lib.check_status\r\npyarrow.lib.ArrowInvalid: Parquet magic bytes not found in footer. Either the file is corrupted or this is not a parquet file.\r\n```\r\nCould it be that I'm using a wrong url, I just copied it from the address bar",
"The URL is wrong indeed, the right one is the one with \"resolve\" (the one you get when clicking on \"download\")- otherwise you try to download an html page ;)\r\n```\r\nhttps://huggingface.co/datasets/bigcode/the-stack-dedup-pjj/resolve/v1.1.a1/data/java/data_0000.parquet\r\n```",
"Ah thanks! So I tried it with the first parquet file and it works, is there a way to know which parquet file was causing the issue since there are a lot of shards?",
"I think you have to try them all :/\r\n\r\nAlternatively you can add a try/catch in `parquet.py` in `datasets` to raise the name of the file that fails at doing `parquet_file = pq.ParquetFile(f)` when you run your initial code\r\n```python\r\nload_dataset(\"bigcode/the-stack-dedup-pjj\", data_dir=\"data/java\", split=\"train\", revision=\"v1.1.a1\", use_auth_token=True)\r\n```\r\nbut it will still iterate on all the files until it fails",
"Ok I will do that",
"I did find the file, and I get the same error as before \r\n```\r\nDownloading data files: 100%|███████████████████| 1/1 [00:00<00:00, 8160.12it/s]\r\nExtracting data files: 100%|████████████████████| 1/1 [00:00<00:00, 1447.81it/s]\r\n \r\n---------------------------------------------------------------------------\r\nArrowInvalid Traceback (most recent call last)\r\nInput In [22], in <cell line: 7>()\r\n 4 data_features = (data[\"train\"].features)\r\n 6 url = \"/home/loubna_huggingface_co/.cache/huggingface/datasets/downloads/93431bc4380de07de8b0ab533666cb5a6120cbe266779e0a63c86bf7717475d7\"\r\n----> 7 data = load_dataset(\"parquet\", \r\n 8 data_files=url,\r\n 9 split=\"train\",\r\n 10 features=data_features,\r\n 11 use_auth_token=True)\r\n\r\nFile /opt/conda/envs/venv/lib/python3.9/site-packages/datasets/load.py:1742, in load_dataset(path, name, data_dir, data_files, split, cache_dir, features, download_config, download_mode, ignore_verifications, keep_in_memory, save_infos, revision, use_auth_token, task, streaming, **config_kwargs)\r\n 1739 try_from_hf_gcs = path not in _PACKAGED_DATASETS_MODULES\r\n 1741 # Download and prepare data\r\n-> 1742 builder_instance.download_and_prepare(\r\n 1743 download_config=download_config,\r\n 1744 download_mode=download_mode,\r\n 1745 ignore_verifications=ignore_verifications,\r\n 1746 try_from_hf_gcs=try_from_hf_gcs,\r\n 1747 use_auth_token=use_auth_token,\r\n 1748 )\r\n 1750 # Build dataset for splits\r\n 1751 keep_in_memory = (\r\n 1752 keep_in_memory if keep_in_memory is not None else is_small_dataset(builder_instance.info.dataset_size)\r\n 1753 )\r\n\r\nFile /opt/conda/envs/venv/lib/python3.9/site-packages/datasets/builder.py:814, in DatasetBuilder.download_and_prepare(self, output_dir, download_config, download_mode, ignore_verifications, try_from_hf_gcs, dl_manager, base_path, use_auth_token, file_format, max_shard_size, storage_options, **download_and_prepare_kwargs)\r\n 808 if not downloaded_from_gcs:\r\n 809 prepare_split_kwargs = {\r\n 810 \"file_format\": file_format,\r\n 811 \"max_shard_size\": max_shard_size,\r\n 812 **download_and_prepare_kwargs,\r\n 813 }\r\n--> 814 self._download_and_prepare(\r\n 815 dl_manager=dl_manager,\r\n 816 verify_infos=verify_infos,\r\n 817 **prepare_split_kwargs,\r\n 818 **download_and_prepare_kwargs,\r\n 819 )\r\n 820 # Sync info\r\n 821 self.info.dataset_size = sum(split.num_bytes for split in self.info.splits.values())\r\n\r\nFile /opt/conda/envs/venv/lib/python3.9/site-packages/datasets/builder.py:905, in DatasetBuilder._download_and_prepare(self, dl_manager, verify_infos, **prepare_split_kwargs)\r\n 901 split_dict.add(split_generator.split_info)\r\n 903 try:\r\n 904 # Prepare split will record examples associated to the split\r\n--> 905 self._prepare_split(split_generator, **prepare_split_kwargs)\r\n 906 except OSError as e:\r\n 907 raise OSError(\r\n 908 \"Cannot find data file. \"\r\n 909 + (self.manual_download_instructions or \"\")\r\n 910 + \"\\nOriginal error:\\n\"\r\n 911 + str(e)\r\n 912 ) from None\r\n\r\nFile /opt/conda/envs/venv/lib/python3.9/site-packages/datasets/builder.py:1502, in ArrowBasedBuilder._prepare_split(self, split_generator, file_format, max_shard_size)\r\n 1500 total_num_examples, total_num_bytes = 0, 0\r\n 1501 try:\r\n-> 1502 for key, table in logging.tqdm(\r\n 1503 generator,\r\n 1504 unit=\" tables\",\r\n 1505 leave=False,\r\n 1506 disable=not logging.is_progress_bar_enabled(),\r\n 1507 ):\r\n 1508 if max_shard_size is not None and writer._num_bytes > max_shard_size:\r\n 1509 num_examples, num_bytes = writer.finalize()\r\n\r\nFile /opt/conda/envs/venv/lib/python3.9/site-packages/tqdm/std.py:1195, in tqdm.__iter__(self)\r\n 1192 time = self._time\r\n 1194 try:\r\n-> 1195 for obj in iterable:\r\n 1196 yield obj\r\n 1197 # Update and possibly print the progressbar.\r\n 1198 # Note: does not call self.update(1) for speed optimisation.\r\n\r\nFile /opt/conda/envs/venv/lib/python3.9/site-packages/datasets/packaged_modules/parquet/parquet.py:67, in Parquet._generate_tables(self, files)\r\n 65 for file_idx, file in enumerate(itertools.chain.from_iterable(files)):\r\n 66 with open(file, \"rb\") as f:\r\n---> 67 parquet_file = pq.ParquetFile(f)\r\n 68 try:\r\n 69 for batch_idx, record_batch in enumerate(\r\n 70 parquet_file.iter_batches(batch_size=self.config.batch_size, columns=self.config.columns)\r\n 71 ):\r\n\r\nFile /opt/conda/envs/venv/lib/python3.9/site-packages/pyarrow/parquet/__init__.py:286, in ParquetFile.__init__(self, source, metadata, common_metadata, read_dictionary, memory_map, buffer_size, pre_buffer, coerce_int96_timestamp_unit, decryption_properties, thrift_string_size_limit, thrift_container_size_limit)\r\n 280 def __init__(self, source, *, metadata=None, common_metadata=None,\r\n 281 read_dictionary=None, memory_map=False, buffer_size=0,\r\n 282 pre_buffer=False, coerce_int96_timestamp_unit=None,\r\n 283 decryption_properties=None, thrift_string_size_limit=None,\r\n 284 thrift_container_size_limit=None):\r\n 285 self.reader = ParquetReader()\r\n--> 286 self.reader.open(\r\n 287 source, use_memory_map=memory_map,\r\n 288 buffer_size=buffer_size, pre_buffer=pre_buffer,\r\n 289 read_dictionary=read_dictionary, metadata=metadata,\r\n 290 coerce_int96_timestamp_unit=coerce_int96_timestamp_unit,\r\n 291 decryption_properties=decryption_properties,\r\n 292 thrift_string_size_limit=thrift_string_size_limit,\r\n 293 thrift_container_size_limit=thrift_container_size_limit,\r\n 294 )\r\n 295 self.common_metadata = common_metadata\r\n 296 self._nested_paths_by_prefix = self._build_nested_paths()\r\n\r\nFile /opt/conda/envs/venv/lib/python3.9/site-packages/pyarrow/_parquet.pyx:1227, in pyarrow._parquet.ParquetReader.open()\r\n\r\nFile /opt/conda/envs/venv/lib/python3.9/site-packages/pyarrow/error.pxi:100, in pyarrow.lib.check_status()\r\n\r\nArrowInvalid: Parquet magic bytes not found in footer. Either the file is corrupted or this is not a parquet file.\r\n```",
"Can you check the JSON file associated to `/home/loubna_huggingface_co/.cache/huggingface/datasets/downloads/93431bc4380de07de8b0ab533666cb5a6120cbe266779e0a63c86bf7717475d7` ? In the JSON file we can know from where it was downloaded\r\n\r\nYou can find it at `/home/loubna_huggingface_co/.cache/huggingface/datasets/downloads/93431bc4380de07de8b0ab533666cb5a6120cbe266779e0a63c86bf7717475d7.json`",
"It's this file `https://huggingface.co/datasets/bigcode/the-stack-dedup-pjj/resolve/f48656daa9f3a3607dacf8b57a65810a6a7a7f73/data/java/data_0022.parquet` loading it gives the same error",
"I'm able to load it properly using\r\n```python\r\nds = load_dataset(\"parquet\", data_files=a_parquet_file_url, use_auth_token=token)\r\n```\r\n\r\nMy guess is that your download was corrupted. Please delete `93431bc4380de07de8b0ab533666cb5a6120cbe266779e0a63c86bf7717475d7` and `93431bc4380de07de8b0ab533666cb5a6120cbe266779e0a63c86bf7717475d7.json` locally and try again",
"That worked, thanks! But I thought if something went wrong with a download `datasets` creates new cache for all the files, that's not the case? (at some point I even changed dataset versions so it was still using that cache?)",
"Cool !\r\n\r\n> But I thought if something went wrong with a download datasets creates new cache for all the files\r\n\r\nWe don't perform integrity verifications if we don't know in advance the hash of the file to download.\r\n\r\n> at some point I even changed dataset versions so it was still using that cache?\r\n\r\n`datasets` caches the files by URL and ETag. If the content of a file changes, then the ETag changes and so it redownloads the file",
"I see, thank you!\r\n",
"I experience the same error in v 2.12.0. But found out it was due to one column from polars was a categorical dtype (related to the error from #5706. Temporarily resolved it by casting the column to str instead."
] |
1,455,252,626
| 5,263
|
Save a dataset in a determined number of shards
|
closed
| 2022-11-18T14:43:54
| 2022-12-14T18:22:59
| 2022-12-14T18:22:59
|
https://github.com/huggingface/datasets/issues/5263
| null |
lhoestq
| false
|
[] |
1,455,171,100
| 5,262
|
AttributeError: 'Value' object has no attribute 'names'
|
closed
| 2022-11-18T13:58:42
| 2022-11-22T10:09:24
| 2022-11-22T10:09:23
|
https://github.com/huggingface/datasets/issues/5262
| null |
emnaboughariou
| false
|
[
"Hi ! It looks like your \"isDif\" column is a Sequence of Value(\"string\"), not a Sequence of ClassLabel.\r\n\r\nYou can convert your Value(\"string\") feature type to a ClassLabel feature type this way:\r\n```python\r\nfrom datasets import ClassLabel, Sequence\r\n\r\n# provide the label_names yourself\r\nlabel_names = [...]\r\n# OR get them from the dataset\r\nlabel_names = sorted(set(label for labels in raw_datasets[\"train\"][\"isDif\"] for label in labels))\r\n\r\n# Cast to ClassLabel\r\nraw_datasets = raw_datasets.cast_column(\"isDif\", Sequence(ClassLabel(names=label_names)))\r\n```\r\n",
"thank you \r\nit works 💯 "
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.