html_url stringlengths 51 51 | comments stringlengths 67 24.7k | title stringlengths 6 280 | body stringlengths 51 36.2k | comment_length int64 16 1.45k | text stringlengths 190 38.3k | embeddings list |
|---|---|---|---|---|---|---|
https://github.com/huggingface/datasets/issues/5317 | 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 arou... | `ImageFolder` performs poorly with large datasets | ### Describe the bug
While testing image dataset creation, I'm seeing significant performance bottlenecks with imagefolders when scanning a directory structure with large number of images.
## Setup
* Nested directories (5 levels deep)
* 3M+ images
* 1 `metadata.jsonl` file
## Performance Degradation Point... | 186 | `ImageFolder` performs poorly with large datasets
### Describe the bug
While testing image dataset creation, I'm seeing significant performance bottlenecks with imagefolders when scanning a directory structure with large number of images.
## Setup
* Nested directories (5 levels deep)
* 3M+ images
* 1 `metada... | [
-0.36105501651763916,
0.1745452880859375,
-0.1266336739063263,
0.4763830602169037,
-0.022154629230499268,
-0.15415151417255402,
0.0810227245092392,
0.21064671874046326,
0.18640030920505524,
0.016278114169836044,
0.1324722021818161,
0.1541631817817688,
-0.42717045545578003,
0.13595077395439... |
https://github.com/huggingface/datasets/issues/5317 | @lhoestq Thanks!
Perhaps 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. | `ImageFolder` performs poorly with large datasets | ### Describe the bug
While testing image dataset creation, I'm seeing significant performance bottlenecks with imagefolders when scanning a directory structure with large number of images.
## Setup
* Nested directories (5 levels deep)
* 3M+ images
* 1 `metadata.jsonl` file
## Performance Degradation Point... | 36 | `ImageFolder` performs poorly with large datasets
### Describe the bug
While testing image dataset creation, I'm seeing significant performance bottlenecks with imagefolders when scanning a directory structure with large number of images.
## Setup
* Nested directories (5 levels deep)
* 3M+ images
* 1 `metada... | [
-0.36105501651763916,
0.1745452880859375,
-0.1266336739063263,
0.4763830602169037,
-0.022154629230499268,
-0.15415151417255402,
0.0810227245092392,
0.21064671874046326,
0.18640030920505524,
0.016278114169836044,
0.1324722021818161,
0.1541631817817688,
-0.42717045545578003,
0.13595077395439... |
https://github.com/huggingface/datasets/issues/5317 | 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! 🙂 | `ImageFolder` performs poorly with large datasets | ### Describe the bug
While testing image dataset creation, I'm seeing significant performance bottlenecks with imagefolders when scanning a directory structure with large number of images.
## Setup
* Nested directories (5 levels deep)
* 3M+ images
* 1 `metadata.jsonl` file
## Performance Degradation Point... | 31 | `ImageFolder` performs poorly with large datasets
### Describe the bug
While testing image dataset creation, I'm seeing significant performance bottlenecks with imagefolders when scanning a directory structure with large number of images.
## Setup
* Nested directories (5 levels deep)
* 3M+ images
* 1 `metada... | [
-0.36105501651763916,
0.1745452880859375,
-0.1266336739063263,
0.4763830602169037,
-0.022154629230499268,
-0.15415151417255402,
0.0810227245092392,
0.21064671874046326,
0.18640030920505524,
0.016278114169836044,
0.1324722021818161,
0.1541631817817688,
-0.42717045545578003,
0.13595077395439... |
https://github.com/huggingface/datasets/issues/5315 | EDIT:
I think in this case, the metadata files (either README or JSON) should not be read (i.e. `self.info.splits` should be None).
One idea:
- I think ideally we should set this behavior when we pass `--save_info` to the CLI `test`
- However, currently, the builder is unaware of this: `save_info` arg is not pas... | Adding new splits to a dataset script with existing old splits info in metadata's `dataset_info` fails | ### Describe the bug
If you first create a custom dataset with a specific set of splits, generate metadata with `datasets-cli test ... --save_info`, then change your script to include more splits, it fails.
That's what happened in https://huggingface.co/datasets/mrdbourke/food_vision_199_classes/discussions/2#6385f... | 57 | Adding new splits to a dataset script with existing old splits info in metadata's `dataset_info` fails
### Describe the bug
If you first create a custom dataset with a specific set of splits, generate metadata with `datasets-cli test ... --save_info`, then change your script to include more splits, it fails.
That... | [
-0.24759887158870697,
0.12104392051696777,
0.013389099389314651,
0.18365627527236938,
0.024452071636915207,
-0.14231444895267487,
0.42856013774871826,
0.3842959403991699,
-0.08495752513408661,
0.025605861097574234,
-0.011631600558757782,
0.33855947852134705,
-0.1291259229183197,
0.27656638... |
https://github.com/huggingface/datasets/issues/5315 | > I think in this case
@albertvillanova You mean in cases when the script was changed?
I suggest that we:
* 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.
* Error here happens before... | Adding new splits to a dataset script with existing old splits info in metadata's `dataset_info` fails | ### Describe the bug
If you first create a custom dataset with a specific set of splits, generate metadata with `datasets-cli test ... --save_info`, then change your script to include more splits, it fails.
That's what happened in https://huggingface.co/datasets/mrdbourke/food_vision_199_classes/discussions/2#6385f... | 142 | Adding new splits to a dataset script with existing old splits info in metadata's `dataset_info` fails
### Describe the bug
If you first create a custom dataset with a specific set of splits, generate metadata with `datasets-cli test ... --save_info`, then change your script to include more splits, it fails.
That... | [
-0.24759887158870697,
0.12104392051696777,
0.013389099389314651,
0.18365627527236938,
0.024452071636915207,
-0.14231444895267487,
0.42856013774871826,
0.3842959403991699,
-0.08495752513408661,
0.025605861097574234,
-0.011631600558757782,
0.33855947852134705,
-0.1291259229183197,
0.27656638... |
https://github.com/huggingface/datasets/issues/5315 | I edited my previous comment:
- First I proposed setting `self.info.splits` to None when `ignore_verifications=True`
- I thought it was the easiest implementation because `ignore_verifications` is passed to `DatasetBuilder.download_and_prepare`
- However, afterwards, I realized this might not be a good idea for ... | Adding new splits to a dataset script with existing old splits info in metadata's `dataset_info` fails | ### Describe the bug
If you first create a custom dataset with a specific set of splits, generate metadata with `datasets-cli test ... --save_info`, then change your script to include more splits, it fails.
That's what happened in https://huggingface.co/datasets/mrdbourke/food_vision_199_classes/discussions/2#6385f... | 201 | Adding new splits to a dataset script with existing old splits info in metadata's `dataset_info` fails
### Describe the bug
If you first create a custom dataset with a specific set of splits, generate metadata with `datasets-cli test ... --save_info`, then change your script to include more splits, it fails.
That... | [
-0.24759887158870697,
0.12104392051696777,
0.013389099389314651,
0.18365627527236938,
0.024452071636915207,
-0.14231444895267487,
0.42856013774871826,
0.3842959403991699,
-0.08495752513408661,
0.025605861097574234,
-0.011631600558757782,
0.33855947852134705,
-0.1291259229183197,
0.27656638... |
https://github.com/huggingface/datasets/issues/5314 | @JonathanAlis also note that the metrics are deprecated in our `datasets` library.
Please, use the new library 🤗 Evaluate instead: https://huggingface.co/docs/evaluate | Datasets: classification_report() got an unexpected keyword argument 'suffix' | https://github.com/huggingface/datasets/blob/main/metrics/seqeval/seqeval.py
> import datasets
predictions = [['O', 'O', 'B-MISC', 'I-MISC', 'I-MISC', 'I-MISC', 'O'], ['B-PER', 'I-PER', 'O']]
references = [['O', 'O', 'O', 'B-MISC', 'I-MISC', 'I-MISC', 'O'], ['B-PER', 'I-PER', 'O']]
seqeval = datasets.load_metri... | 21 | Datasets: classification_report() got an unexpected keyword argument 'suffix'
https://github.com/huggingface/datasets/blob/main/metrics/seqeval/seqeval.py
> import datasets
predictions = [['O', 'O', 'B-MISC', 'I-MISC', 'I-MISC', 'I-MISC', 'O'], ['B-PER', 'I-PER', 'O']]
references = [['O', 'O', 'O', 'B-MISC', '... | [
-0.08563310652971268,
-0.19035319983959198,
-0.01853889971971512,
0.18329522013664246,
0.2619512677192688,
0.17882893979549408,
0.055492956191301346,
0.25525185465812683,
-0.1317923367023468,
-0.06356767565011978,
-0.26776018738746643,
0.306918203830719,
-0.1053706556558609,
-0.00352755188... |
https://github.com/huggingface/datasets/issues/5305 | Thanks for reporting @JoelNiklaus.
Please 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/discussio... | Dataset joelito/mc4_legal does not work with multiple files | ### Describe the bug
The dataset https://huggingface.co/datasets/joelito/mc4_legal works for languages like bg with a single data file, but not for languages with multiple files like de. It shows zero rows for the de dataset.
joelniklaus@Joels-MacBook-Pro ~/N/P/C/L/p/m/mc4_legal (main) [1]> python test_mc4_legal.... | 54 | Dataset joelito/mc4_legal does not work with multiple files
### Describe the bug
The dataset https://huggingface.co/datasets/joelito/mc4_legal works for languages like bg with a single data file, but not for languages with multiple files like de. It shows zero rows for the de dataset.
joelniklaus@Joels-MacBook-... | [
-0.1884015053510666,
-0.22016529738903046,
-0.04741975665092468,
0.4664088487625122,
0.2048230767250061,
0.09562930464744568,
0.5133573412895203,
0.22911669313907623,
-0.21484509110450745,
0.021942030638456345,
0.08973903954029083,
-0.1431405395269394,
-0.2039322555065155,
0.13756760954856... |
https://github.com/huggingface/datasets/issues/5304 | 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:
```python
wav_paths = sorted(Path(data_dir).glob(f"**/{split}/**/*.wav"))
wav_paths = wav_paths if wav_paths else sorted(Path(data_dir).glob(f"**/{split.upper()}/**/... | timit_asr doesn't load the test split. | ### Describe the bug
When I use the function ```timit = load_dataset('timit_asr', data_dir=data_dir)```, it only loads train split, not test split.
I tried to change the directory and filename to lower case to upper case for the test split, but it does not work at all.
```python
DatasetDict({
train: Datase... | 42 | timit_asr doesn't load the test split.
### Describe the bug
When I use the function ```timit = load_dataset('timit_asr', data_dir=data_dir)```, it only loads train split, not test split.
I tried to change the directory and filename to lower case to upper case for the test split, but it does not work at all.
``... | [
-0.24493855237960815,
-0.11158931255340576,
-0.020474031567573547,
0.38997671008110046,
0.19437098503112793,
0.03327915817499161,
0.4489307403564453,
0.22222363948822021,
-0.21772345900535583,
0.05952703580260277,
-0.22904865443706512,
0.2974380850791931,
-0.18306168913841248,
0.3608002662... |
https://github.com/huggingface/datasets/issues/5295 | 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). | Extractions failed when .zip file located on read-only path (e.g., SageMaker FastFile mode) | ### Describe the bug
Hi,
`load_dataset()` does not work .zip files located on a read-only directory. Looks like it's because Dataset creates a lock file in the [same directory](https://github.com/huggingface/datasets/blob/df4bdd365f2abb695f113cbf8856a925bc70901b/src/datasets/utils/extract.py) as the .zip file.
... | 27 | Extractions failed when .zip file located on read-only path (e.g., SageMaker FastFile mode)
### Describe the bug
Hi,
`load_dataset()` does not work .zip files located on a read-only directory. Looks like it's because Dataset creates a lock file in the [same directory](https://github.com/huggingface/datasets/blo... | [
-0.19462928175926208,
-0.12467707693576813,
0.03719492256641388,
0.4657360911369324,
0.08938072621822357,
-0.10493163764476776,
0.30083125829696655,
0.15286733210086823,
0.026789501309394836,
0.1481025516986847,
-0.434200257062912,
0.42163169384002686,
-0.09773548692464828,
0.3973968923091... |
https://github.com/huggingface/datasets/issues/5295 | 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 | Extractions failed when .zip file located on read-only path (e.g., SageMaker FastFile mode) | ### Describe the bug
Hi,
`load_dataset()` does not work .zip files located on a read-only directory. Looks like it's because Dataset creates a lock file in the [same directory](https://github.com/huggingface/datasets/blob/df4bdd365f2abb695f113cbf8856a925bc70901b/src/datasets/utils/extract.py) as the .zip file.
... | 29 | Extractions failed when .zip file located on read-only path (e.g., SageMaker FastFile mode)
### Describe the bug
Hi,
`load_dataset()` does not work .zip files located on a read-only directory. Looks like it's because Dataset creates a lock file in the [same directory](https://github.com/huggingface/datasets/blo... | [
-0.19462928175926208,
-0.12467707693576813,
0.03719492256641388,
0.4657360911369324,
0.08938072621822357,
-0.10493163764476776,
0.30083125829696655,
0.15286733210086823,
0.026789501309394836,
0.1481025516986847,
-0.434200257062912,
0.42163169384002686,
-0.09773548692464828,
0.3973968923091... |
https://github.com/huggingface/datasets/issues/5292 | - Build docs for 2.6.2:
- Commit: a6a5a1cf4cdf1e0be65168aed5a327f543001fe8
- Build docs GH Action: https://github.com/huggingface/datasets/actions/runs/3539470622/jobs/5941404044
- Build docs for 2.7.1:
- Commit: 5ef1ab1cc06c2b7a574bf2df454cd9fcb071ccb2
- Build docs GH Action: https://github.com/huggingfac... | Missing documentation build for versions 2.7.1 and 2.6.2 | After the patch releases [2.7.1](https://github.com/huggingface/datasets/releases/tag/2.7.1) and [2.6.2](https://github.com/huggingface/datasets/releases/tag/2.6.2), the online docs were not properly built (the build_documentation workflow was not triggered).
There was a fix by:
- #5291
However, both documentati... | 28 | Missing documentation build for versions 2.7.1 and 2.6.2
After the patch releases [2.7.1](https://github.com/huggingface/datasets/releases/tag/2.7.1) and [2.6.2](https://github.com/huggingface/datasets/releases/tag/2.6.2), the online docs were not properly built (the build_documentation workflow was not triggered).
... | [
-0.2693377435207367,
-0.29634004831314087,
-0.11461298167705536,
0.015760298818349838,
0.11522068083286285,
-0.11685895174741745,
0.032839085906744,
0.35888347029685974,
-0.0780351459980011,
0.015483304858207703,
0.04930455982685089,
0.21430905163288116,
0.14360952377319336,
0.212121099233... |
https://github.com/huggingface/datasets/issues/5288 | 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. | Lossy json serialization - deserialization of dataset info | ### Describe the bug
Saving a dataset to disk as json (using `to_json`) and then loading it again (using `load_dataset`) results in features whose labels are not type-cast correctly. In the code snippet below, `features.label` should have a label of type `ClassLabel` but has type `Value` instead.
### Steps to re... | 45 | Lossy json serialization - deserialization of dataset info
### Describe the bug
Saving a dataset to disk as json (using `to_json`) and then loading it again (using `load_dataset`) results in features whose labels are not type-cast correctly. In the code snippet below, `features.label` should have a label of type ... | [
0.005198776721954346,
0.21891941130161285,
-0.008309319615364075,
0.42117375135421753,
0.5208909511566162,
0.14566728472709656,
0.4419674873352051,
0.28813275694847107,
-0.20148900151252747,
-0.11444413661956787,
-0.14278250932693481,
0.7578495144844055,
0.030339006334543228,
0.12828744947... |
https://github.com/huggingface/datasets/issues/5286 | I found a solution
If you specifically install datasets==1.18 and then run
import datasets
wiki = datasets.load_dataset('wikipedia', '20200501.en')
then this should work (it worked for me.) | FileNotFoundError: Couldn't find file at https://dumps.wikimedia.org/enwiki/20220301/dumpstatus.json | ### Describe the bug
I follow the steps provided on the website [https://huggingface.co/datasets/wikipedia](https://huggingface.co/datasets/wikipedia)
$ pip install apache_beam mwparserfromhell
>>> from datasets import load_dataset
>>> load_dataset("wikipedia", "20220301.en")
however this results in the follo... | 26 | FileNotFoundError: Couldn't find file at https://dumps.wikimedia.org/enwiki/20220301/dumpstatus.json
### Describe the bug
I follow the steps provided on the website [https://huggingface.co/datasets/wikipedia](https://huggingface.co/datasets/wikipedia)
$ pip install apache_beam mwparserfromhell
>>> from datasets... | [
-0.23028485476970673,
0.1827666163444519,
-0.009024474769830704,
0.26874232292175293,
0.10095718502998352,
0.13449442386627197,
0.10901640355587006,
0.34828805923461914,
0.1558169722557068,
0.21776776015758514,
0.2789648771286011,
0.14793352782726288,
-0.10855913162231445,
-0.3582846522331... |
https://github.com/huggingface/datasets/issues/5284 | Thanks @lhoestq and @alvarobartt!
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!
_c.f._ https://twitter.com/sanchitgandhi99/status/15921... | Features of IterableDataset set to None by remove column | ### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librispeech_asr", "clean", split="validation", streaming=True)
... | 43 | Features of IterableDataset set to None by remove column
### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librisp... | [
-0.261094331741333,
-0.2930840849876404,
-0.019567906856536865,
0.13754194974899292,
0.290933221578598,
-0.026313021779060364,
0.4993182420730591,
0.15073545277118683,
-0.12898756563663483,
0.19063225388526917,
-0.24397878348827362,
0.44379934668540955,
-0.07863340526819229,
-0.04678307101... |
https://github.com/huggingface/datasets/issues/5284 | > Thanks @lhoestq and @alvarobartt!
>
>
>
> 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!
>
>
>
> _c.f._ https://twitter.com/sanchitgan... | Features of IterableDataset set to None by remove column | ### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librispeech_asr", "clean", split="validation", streaming=True)
... | 100 | Features of IterableDataset set to None by remove column
### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librisp... | [
-0.261094331741333,
-0.2930840849876404,
-0.019567906856536865,
0.13754194974899292,
0.290933221578598,
-0.026313021779060364,
0.4993182420730591,
0.15073545277118683,
-0.12898756563663483,
0.19063225388526917,
-0.24397878348827362,
0.44379934668540955,
-0.07863340526819229,
-0.04678307101... |
https://github.com/huggingface/datasets/issues/5284 | @sanchit-gandhi PR is ready and open for review at #5287, but there's still one issue I may need @lhoestq's input :hugs: | Features of IterableDataset set to None by remove column | ### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librispeech_asr", "clean", split="validation", streaming=True)
... | 21 | Features of IterableDataset set to None by remove column
### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librisp... | [
-0.261094331741333,
-0.2930840849876404,
-0.019567906856536865,
0.13754194974899292,
0.290933221578598,
-0.026313021779060364,
0.4993182420730591,
0.15073545277118683,
-0.12898756563663483,
0.19063225388526917,
-0.24397878348827362,
0.44379934668540955,
-0.07863340526819229,
-0.04678307101... |
https://github.com/huggingface/datasets/issues/5284 | Let us know @sanchit-gandhi if you need a new release of `datasets` soon with this fix included :) | Features of IterableDataset set to None by remove column | ### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librispeech_asr", "clean", split="validation", streaming=True)
... | 18 | Features of IterableDataset set to None by remove column
### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librisp... | [
-0.261094331741333,
-0.2930840849876404,
-0.019567906856536865,
0.13754194974899292,
0.290933221578598,
-0.026313021779060364,
0.4993182420730591,
0.15073545277118683,
-0.12898756563663483,
0.19063225388526917,
-0.24397878348827362,
0.44379934668540955,
-0.07863340526819229,
-0.04678307101... |
https://github.com/huggingface/datasets/issues/5284 | Thanks for the fix guys! We can direct people to install `datasets` from main if that's easier! | Features of IterableDataset set to None by remove column | ### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librispeech_asr", "clean", split="validation", streaming=True)
... | 17 | Features of IterableDataset set to None by remove column
### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librisp... | [
-0.261094331741333,
-0.2930840849876404,
-0.019567906856536865,
0.13754194974899292,
0.290933221578598,
-0.026313021779060364,
0.4993182420730591,
0.15073545277118683,
-0.12898756563663483,
0.19063225388526917,
-0.24397878348827362,
0.44379934668540955,
-0.07863340526819229,
-0.04678307101... |
https://github.com/huggingface/datasets/issues/5284 | 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/datas... | Features of IterableDataset set to None by remove column | ### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librispeech_asr", "clean", split="validation", streaming=True)
... | 47 | Features of IterableDataset set to None by remove column
### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librisp... | [
-0.261094331741333,
-0.2930840849876404,
-0.019567906856536865,
0.13754194974899292,
0.290933221578598,
-0.026313021779060364,
0.4993182420730591,
0.15073545277118683,
-0.12898756563663483,
0.19063225388526917,
-0.24397878348827362,
0.44379934668540955,
-0.07863340526819229,
-0.04678307101... |
https://github.com/huggingface/datasets/issues/5284 | Still facing the same issue though:
```
from datasets import IterableDatasetDict, load_dataset
raw_datasets = vectorized_datasets = IterableDatasetDict()
raw_datasets["train"] = load_dataset("asennoussi/private", split="train", use_auth_token=True, streaming=True)
raw_datasets["test"] = load_dataset("asenno... | Features of IterableDataset set to None by remove column | ### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librispeech_asr", "clean", split="validation", streaming=True)
... | 120 | Features of IterableDataset set to None by remove column
### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librisp... | [
-0.261094331741333,
-0.2930840849876404,
-0.019567906856536865,
0.13754194974899292,
0.290933221578598,
-0.026313021779060364,
0.4993182420730591,
0.15073545277118683,
-0.12898756563663483,
0.19063225388526917,
-0.24397878348827362,
0.44379934668540955,
-0.07863340526819229,
-0.04678307101... |
https://github.com/huggingface/datasets/issues/5284 | Hmm weird, could you try to print
```python
print("Processed features: ", vectorized_datasets['train'].features)
```
again after iterating over the `vectorized_datasets`? In the code above, should be last line :) | Features of IterableDataset set to None by remove column | ### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librispeech_asr", "clean", split="validation", streaming=True)
... | 28 | Features of IterableDataset set to None by remove column
### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librisp... | [
-0.261094331741333,
-0.2930840849876404,
-0.019567906856536865,
0.13754194974899292,
0.290933221578598,
-0.026313021779060364,
0.4993182420730591,
0.15073545277118683,
-0.12898756563663483,
0.19063225388526917,
-0.24397878348827362,
0.44379934668540955,
-0.07863340526819229,
-0.04678307101... |
https://github.com/huggingface/datasets/issues/5284 | Didn't seem to fix it:
```
Original features: dict_keys(['path', 'audio', 'sentence'])
Processed features: None
Processed features: None
``` | Features of IterableDataset set to None by remove column | ### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librispeech_asr", "clean", split="validation", streaming=True)
... | 18 | Features of IterableDataset set to None by remove column
### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librisp... | [
-0.261094331741333,
-0.2930840849876404,
-0.019567906856536865,
0.13754194974899292,
0.290933221578598,
-0.026313021779060364,
0.4993182420730591,
0.15073545277118683,
-0.12898756563663483,
0.19063225388526917,
-0.24397878348827362,
0.44379934668540955,
-0.07863340526819229,
-0.04678307101... |
https://github.com/huggingface/datasets/issues/5284 | Actually the culprit looks to be this one:
`vectorized_datasets = raw_datasets.map(prepare_dataset).with_format("torch")`
When I remove this line: `vectorized_datasets = vectorized_datasets.remove_columns(['input_length', 'labels_length']+list(next(iter(raw_datasets.values())).features))`
I still get
```
Proce... | Features of IterableDataset set to None by remove column | ### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librispeech_asr", "clean", split="validation", streaming=True)
... | 28 | Features of IterableDataset set to None by remove column
### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librisp... | [
-0.261094331741333,
-0.2930840849876404,
-0.019567906856536865,
0.13754194974899292,
0.290933221578598,
-0.026313021779060364,
0.4993182420730591,
0.15073545277118683,
-0.12898756563663483,
0.19063225388526917,
-0.24397878348827362,
0.44379934668540955,
-0.07863340526819229,
-0.04678307101... |
https://github.com/huggingface/datasets/issues/5284 | > The culprit is definitely `.map` Just validated it. Any idea please?
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 https://github.com/huggingface/datasets/issues/3888
Anyway, now you can pass the `features`... | Features of IterableDataset set to None by remove column | ### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librispeech_asr", "clean", split="validation", streaming=True)
... | 101 | Features of IterableDataset set to None by remove column
### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librisp... | [
-0.261094331741333,
-0.2930840849876404,
-0.019567906856536865,
0.13754194974899292,
0.290933221578598,
-0.026313021779060364,
0.4993182420730591,
0.15073545277118683,
-0.12898756563663483,
0.19063225388526917,
-0.24397878348827362,
0.44379934668540955,
-0.07863340526819229,
-0.04678307101... |
https://github.com/huggingface/datasets/issues/5284 | @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 abo... | Features of IterableDataset set to None by remove column | ### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librispeech_asr", "clean", split="validation", streaming=True)
... | 56 | Features of IterableDataset set to None by remove column
### Describe the bug
The `remove_column` method of the IterableDataset sets the dataset features to None.
### Steps to reproduce the bug
```python
from datasets import Audio, load_dataset
# load LS in streaming mode
dataset = load_dataset("librisp... | [
-0.261094331741333,
-0.2930840849876404,
-0.019567906856536865,
0.13754194974899292,
0.290933221578598,
-0.026313021779060364,
0.4993182420730591,
0.15073545277118683,
-0.12898756563663483,
0.19063225388526917,
-0.24397878348827362,
0.44379934668540955,
-0.07863340526819229,
-0.04678307101... |
https://github.com/huggingface/datasets/issues/5281 | 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... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 28 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.17933502793312073,
0.11833947896957397,
-0.133915975689888,
0.07898159325122833,
0.12148794531822205,
-0.06233625486493111,
0.2835327684879303,
0.1423354297876358,
0.29781562089920044,
0.13451774418354034,
-0.06672105193138123,
0.17026948928833008,
-0.22052764892578125,
0.80269438028335... |
https://github.com/huggingface/datasets/issues/5281 | +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. | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 36 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.27272918820381165,
-0.034030571579933167,
-0.13250580430030823,
0.08983898907899857,
0.2238250970840454,
0.01593600958585739,
0.24261833727359772,
-0.08133567869663239,
0.1618099957704544,
0.07210321724414825,
-0.19991600513458252,
0.29661497473716736,
-0.3891263008117676,
0.82465410232... |
https://github.com/huggingface/datasets/issues/5281 | Adding to the conversation, Dask also uses `fsspec` for this feature.
[Dask: How to connect to remote data](https://docs.dask.org/en/stable/how-to/connect-to-remote-data.html)
Happy to help on this feature :D | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 25 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.23842087388038635,
-0.1173187717795372,
-0.12805670499801636,
0.03234292194247246,
0.17283374071121216,
-0.04650694131851196,
0.19936655461788177,
0.04815227910876274,
0.21960331499576569,
0.08897990733385086,
-0.08867645263671875,
0.1969701051712036,
-0.16965420544147491,
0.88736957311... |
https://github.com/huggingface/datasets/issues/5281 | +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.
Currently we could use `builder.download_a... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 78 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.32151031494140625,
0.11817923188209534,
-0.10677763819694519,
0.13897572457790375,
0.1547628939151764,
0.09225157648324966,
0.18456090986728668,
0.014310022816061974,
0.2103939950466156,
-0.0009158328175544739,
0.04465002566576004,
0.19866116344928741,
-0.2637074291706085,
0.65082997083... |
https://github.com/huggingface/datasets/issues/5281 | 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`. | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 39 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.2028985321521759,
-0.14597223699092865,
-0.14027895033359528,
0.05473877862095833,
0.2059151828289032,
-0.03967658430337906,
0.2350216805934906,
0.027860045433044434,
0.26465243101119995,
0.07531090080738068,
-0.15618765354156494,
0.24585089087486267,
-0.19070744514465332,
0.83104974031... |
https://github.com/huggingface/datasets/issues/5281 | It seems currently the `cached_path` function handles all URLs by `get_from_cache` that only supports `ftp` and `http(s)` here:
https://github.com/huggingface/datasets/blob/b5672a956d5de864e6f5550e493527d962d6ae55/src/datasets/utils/file_utils.py#L181
I guess one can add another condition that handles `s3://` or `g... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 35 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.1167890802025795,
0.03152339160442352,
-0.11542992293834686,
0.09765052795410156,
0.17531056702136993,
-0.057010818272829056,
0.08391191810369492,
0.07078969478607178,
0.24949762225151062,
-0.13037090003490448,
-0.05985070765018463,
0.10622628033161163,
-0.2772143483161926,
0.6344969868... |
https://github.com/huggingface/datasets/issues/5281 | I could use this functionality, so I put together a PR using @kyamagu's suggestion to use `fsspec` in `datasets.utils.file_utils`
https://github.com/huggingface/datasets/pull/5580 | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 20 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.15794751048088074,
-0.09473351389169693,
-0.14925864338874817,
0.10644612461328506,
0.2577477693557739,
0.001848362386226654,
0.15471938252449036,
-0.02633676491677761,
0.1780385971069336,
0.019037984311580658,
-0.05524980276823044,
0.23475602269172668,
-0.265080988407135,
0.80302298069... |
https://github.com/huggingface/datasets/issues/5281 | Thanks @dwyatte for adding support for fsspec urls
Let me just reopen this since the original issue is not resolved | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 20 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.12578171491622925,
-0.09776636213064194,
-0.12361013889312744,
0.10396529734134674,
0.2162245512008667,
-0.03363677114248276,
0.15729855000972748,
-0.06592381000518799,
0.1577867567539215,
0.009724512696266174,
-0.07180546224117279,
0.18582415580749512,
-0.20637187361717224,
0.817408800... |
https://github.com/huggingface/datasets/issues/5281 | 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! | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 21 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.09307672083377838,
-0.14621078968048096,
-0.12993310391902924,
0.18423116207122803,
0.1994064748287201,
0.01147422194480896,
0.2190307080745697,
-0.1165558397769928,
0.13301369547843933,
-0.01619921624660492,
-0.012743860483169556,
0.14910396933555603,
-0.23013703525066376,
0.8589407205... |
https://github.com/huggingface/datasets/issues/5281 | 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 | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 43 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.10912920534610748,
-0.17285233736038208,
-0.0897660180926323,
0.05335089936852455,
0.22264856100082397,
-0.09287696331739426,
0.27672240138053894,
0.06577986478805542,
0.1796685755252838,
0.0005553141236305237,
-0.23923957347869873,
0.21224790811538696,
-0.14371603727340698,
0.667562544... |
https://github.com/huggingface/datasets/issues/5281 | 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:
InvalidSchema: No connection adapters were found for 's3://bucket/folder/'. I'... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 92 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.2100405991077423,
0.08720414340496063,
-0.052655015140771866,
0.142103910446167,
0.11030682176351547,
-0.054292500019073486,
0.27416661381721497,
-0.006058541126549244,
0.0784444808959961,
-0.024943456053733826,
-0.19915415346622467,
0.1984845995903015,
-0.01397683471441269,
0.808444857... |
https://github.com/huggingface/datasets/issues/5281 | 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`
Note that while testing this just now, I did note a discrepancy between gcsfs an... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 130 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.22709104418754578,
0.10611672699451447,
-0.09424984455108643,
0.04079977422952652,
0.2034299075603485,
-0.00865822285413742,
0.43176451325416565,
-0.1008283942937851,
0.18912360072135925,
0.0013836249709129333,
-0.11103907227516174,
0.2538589537143707,
-0.08419495820999146,
0.7252944707... |
https://github.com/huggingface/datasets/issues/5281 | > 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 aio... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 109 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.11338277906179428,
-0.0041427165269851685,
-0.12317997962236404,
-0.020559553056955338,
0.18401798605918884,
-0.10166347026824951,
0.4652293920516968,
-0.13734617829322815,
0.13873903453350067,
0.11933740973472595,
0.0024716705083847046,
0.1991029530763626,
-0.03207051381468773,
0.71764... |
https://github.com/huggingface/datasets/issues/5281 | @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... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 274 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.20576931536197662,
-0.1730315387248993,
-0.1266249120235443,
0.025293510407209396,
0.10306775569915771,
-0.10649237781763077,
0.3078857660293579,
-0.11788011342287064,
0.0971309095621109,
-0.047364525496959686,
-0.16153864562511444,
0.07935158163309097,
-0.2854681611061096,
0.8079638481... |
https://github.com/huggingface/datasets/issues/5281 | @lhoestq
hello, i still have problem with loading json from S3:
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)
and it throws an error:
TypeError: AioSessio... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 49 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.07479547709226608,
-0.1079692393541336,
-0.08909490704536438,
0.17212776839733124,
0.16700847446918488,
-0.03813118487596512,
0.1875782310962677,
-0.007055429741740227,
0.2004779875278473,
-0.003200612962245941,
-0.08756573498249054,
0.25158366560935974,
-0.11807522177696228,
0.75721287... |
https://github.com/huggingface/datasets/issues/5281 | 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`.
Now `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 expl... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 140 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.1654491424560547,
0.19363754987716675,
-0.1447126865386963,
0.17066946625709534,
0.19591404497623444,
-0.038630224764347076,
0.3390748202800751,
0.06812532246112823,
0.06344951689243317,
0.05108793452382088,
-0.07675494253635406,
0.15149861574172974,
-0.33743005990982056,
0.747261404991... |
https://github.com/huggingface/datasets/issues/5281 | > 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 (se... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 104 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.12195287644863129,
0.12658941745758057,
-0.14086265861988068,
0.26538461446762085,
0.2808706760406494,
0.06839892268180847,
0.22423073649406433,
0.013140494003891945,
0.2562103867530823,
0.12566730380058289,
-0.03907795995473862,
0.22973862290382385,
-0.34713295102119446,
0.611405611038... |
https://github.com/huggingface/datasets/issues/5281 | I'm seeing same problem as @hjq133 with following versions:
```
datasets==2.15.0
(venv) ➜ finetuning-llama2 git:(main) ✗ pip freeze | grep s3fs
s3fs==2023.10.0
(venv) ➜ finetuning-llama2 git:(main) ✗ pip freeze | grep fsspec
fsspec==2023.10.0
``` | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 34 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.18196001648902893,
-0.02283702790737152,
-0.11067073047161102,
0.12068088352680206,
0.24369272589683533,
-0.06892702728509903,
0.2155008465051651,
0.010046415030956268,
0.16540087759494781,
0.091794453561306,
-0.08870047330856323,
0.3211115002632141,
-0.25414663553237915,
0.733625173568... |
https://github.com/huggingface/datasets/issues/5281 | > @lhoestq hello, i still have problem with loading json from S3:
>
> 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)
>
> and it throws an error: TypeError: AioSession.**init**() ... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 83 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
0.000037923455238342285,
-0.019027993083000183,
-0.07329682260751724,
0.26487356424331665,
0.07412988692522049,
-0.09869390726089478,
0.28396642208099365,
-0.09188677370548248,
0.1530681848526001,
0.050795778632164,
-0.14965775609016418,
0.23283550143241882,
-0.05071890354156494,
0.6841099... |
https://github.com/huggingface/datasets/issues/5281 | 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 tha... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 66 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.2460056096315384,
-0.1542784571647644,
-0.04496544972062111,
0.03817639499902725,
0.33450695872306824,
0.0012636780738830566,
0.22305068373680115,
0.08378459513187408,
0.1938738375902176,
0.05832371488213539,
-0.18344515562057495,
0.17693781852722168,
-0.21696124970912933,
0.65147995948... |
https://github.com/huggingface/datasets/issues/5281 | @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_datase... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 92 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.15654803812503815,
0.009563252329826355,
-0.11528217792510986,
0.011855952441692352,
0.04420935735106468,
-0.08809714019298553,
0.24136805534362793,
-0.0680532157421112,
0.08408168703317642,
-0.008036598563194275,
-0.05647020787000656,
0.26827874779701233,
-0.24570590257644653,
0.800529... |
https://github.com/huggingface/datasets/issues/5281 | @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
It looks like `datasets` uses a default `DownloadConfig` which is where some potentially unintended storage options are gett... | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 157 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.0414397194981575,
-0.10600695013999939,
-0.07206211984157562,
0.10603192448616028,
0.2360774129629135,
-0.17956854403018951,
0.24351853132247925,
-0.1231684684753418,
0.058073267340660095,
-0.008839428424835205,
0.03298545628786087,
0.28888067603111267,
-0.21818920969963074,
0.756968915... |
https://github.com/huggingface/datasets/issues/5281 | 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 | Support cloud storage in load_dataset | Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` and `save_to_disk`.
This has been reque... | 18 | Support cloud storage in load_dataset
Would be nice to be able to do
```python
data_files=["s3://..."] # or gs:// or any cloud storage path
storage_options = {...}
load_dataset(..., data_files=data_files, storage_options=storage_options)
```
The idea would be to use `fsspec` as in `download_and_prepare` an... | [
-0.17138615250587463,
-0.12152993679046631,
-0.13658443093299866,
0.06388713419437408,
0.21852923929691315,
-0.003932282328605652,
0.19721408188343048,
-0.009359964169561863,
0.22520838677883148,
0.05114608258008957,
-0.12447507679462433,
0.19488292932510376,
-0.20609323680400848,
0.806365... |
https://github.com/huggingface/datasets/issues/5280 | Hi,
3.8.13
Get Outlook for Android<https://aka.ms/AAb9ysg>
________________________________
From: Quentin Lhoest ***@***.***>
Sent: Tuesday, November 22, 2022 2:37:02 PM
To: huggingface/datasets ***@***.***>
Cc: feketedavid1012 ***@***.***>; Author ***@***.***>
Subject: Re: [huggingface/datasets] Import error (Issue ... | Import error | https://github.com/huggingface/datasets/blob/cd3d8e637cfab62d352a3f4e5e60e96597b5f0e9/src/datasets/__init__.py#L28
Hy,
I have error at the above line. I have python version 3.8.13, the message says I need python>=3.7, which is True, but I think the if statement not working properly (or the message wrong) | 69 | Import error
https://github.com/huggingface/datasets/blob/cd3d8e637cfab62d352a3f4e5e60e96597b5f0e9/src/datasets/__init__.py#L28
Hy,
I have error at the above line. I have python version 3.8.13, the message says I need python>=3.7, which is True, but I think the if statement not working properly (or the message w... | [
-0.09654354304075241,
-0.7863255739212036,
-0.11001820862293243,
0.42073991894721985,
0.2069012075662613,
0.12516063451766968,
-0.06425055861473083,
0.26119884848594666,
0.25540074706077576,
0.06383652985095978,
0.0825561136007309,
0.01027308776974678,
-0.1894727647304535,
0.13995850086212... |
https://github.com/huggingface/datasets/issues/5280 | Then it should work as expected if you use the same python when using `datasets`
Please make sure you're running your code in the right environment | Import error | https://github.com/huggingface/datasets/blob/cd3d8e637cfab62d352a3f4e5e60e96597b5f0e9/src/datasets/__init__.py#L28
Hy,
I have error at the above line. I have python version 3.8.13, the message says I need python>=3.7, which is True, but I think the if statement not working properly (or the message wrong) | 26 | Import error
https://github.com/huggingface/datasets/blob/cd3d8e637cfab62d352a3f4e5e60e96597b5f0e9/src/datasets/__init__.py#L28
Hy,
I have error at the above line. I have python version 3.8.13, the message says I need python>=3.7, which is True, but I think the if statement not working properly (or the message w... | [
-0.2534981966018677,
-0.437334805727005,
-0.18320974707603455,
0.3004244565963745,
0.24939267337322235,
0.22554433345794678,
0.032267022877931595,
0.45042136311531067,
0.23203495144844055,
0.024222321808338165,
-0.045096416026353836,
-0.061573926359415054,
-0.18736889958381653,
0.336300313... |
https://github.com/huggingface/datasets/issues/5280 | It's the right environment. But in if statement I have
"3.8.13" < 3.7
And 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 statemen... | Import error | https://github.com/huggingface/datasets/blob/cd3d8e637cfab62d352a3f4e5e60e96597b5f0e9/src/datasets/__init__.py#L28
Hy,
I have error at the above line. I have python version 3.8.13, the message says I need python>=3.7, which is True, but I think the if statement not working properly (or the message wrong) | 147 | Import error
https://github.com/huggingface/datasets/blob/cd3d8e637cfab62d352a3f4e5e60e96597b5f0e9/src/datasets/__init__.py#L28
Hy,
I have error at the above line. I have python version 3.8.13, the message says I need python>=3.7, which is True, but I think the if statement not working properly (or the message w... | [
-0.047881655395030975,
-0.7668336629867554,
-0.12649723887443542,
0.33807098865509033,
0.16299641132354736,
0.03349854052066803,
-0.06797115504741669,
0.2466142177581787,
0.2913607954978943,
-0.02438081055879593,
-0.05494983494281769,
-0.1026596873998642,
-0.063369020819664,
0.218904256820... |
https://github.com/huggingface/datasets/issues/5278 | Can you try to remove "drop_labels=false" ? It may force the loader to infer the labels instead of reading the metadata | load_dataset does not read jsonl metadata file properly | ### Describe the bug
Hi, I'm following [this page](https://huggingface.co/docs/datasets/image_dataset) to create a dataset of images and captions via an image folder and a metadata.json file, but I can't seem to get the dataloader to recognize the "text" column. It just spits out "image" and "label" as features.
B... | 21 | load_dataset does not read jsonl metadata file properly
### Describe the bug
Hi, I'm following [this page](https://huggingface.co/docs/datasets/image_dataset) to create a dataset of images and captions via an image folder and a metadata.json file, but I can't seem to get the dataloader to recognize the "text" column... | [
0.00862911343574524,
0.29756903648376465,
0.0013801641762256622,
0.8008169531822205,
0.39346379041671753,
0.05997225642204285,
0.39974090456962585,
0.21833279728889465,
0.06595367938280106,
-0.0907503217458725,
-0.08651769161224365,
0.565157413482666,
-0.14890941977500916,
0.12042210251092... |
https://github.com/huggingface/datasets/issues/5278 | 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 | load_dataset does not read jsonl metadata file properly | ### Describe the bug
Hi, I'm following [this page](https://huggingface.co/docs/datasets/image_dataset) to create a dataset of images and captions via an image folder and a metadata.json file, but I can't seem to get the dataloader to recognize the "text" column. It just spits out "image" and "label" as features.
B... | 33 | load_dataset does not read jsonl metadata file properly
### Describe the bug
Hi, I'm following [this page](https://huggingface.co/docs/datasets/image_dataset) to create a dataset of images and captions via an image folder and a metadata.json file, but I can't seem to get the dataloader to recognize the "text" column... | [
0.009904853999614716,
0.3267070949077606,
0.01878252997994423,
0.8210528492927551,
0.37091687321662903,
0.02410455048084259,
0.42159709334373474,
0.2379225492477417,
0.018689006567001343,
-0.1290813684463501,
-0.09945640712976456,
0.544184148311615,
-0.15521135926246643,
0.1085872799158096... |
https://github.com/huggingface/datasets/issues/5278 | > Can you try updating `datasets` ? Metadata support was added in `datasets` 2.4
Update: This was the issue. | load_dataset does not read jsonl metadata file properly | ### Describe the bug
Hi, I'm following [this page](https://huggingface.co/docs/datasets/image_dataset) to create a dataset of images and captions via an image folder and a metadata.json file, but I can't seem to get the dataloader to recognize the "text" column. It just spits out "image" and "label" as features.
B... | 19 | load_dataset does not read jsonl metadata file properly
### Describe the bug
Hi, I'm following [this page](https://huggingface.co/docs/datasets/image_dataset) to create a dataset of images and captions via an image folder and a metadata.json file, but I can't seem to get the dataloader to recognize the "text" column... | [
-0.006252601742744446,
0.3133275508880615,
0.008775286376476288,
0.778401255607605,
0.385869562625885,
0.0617540180683136,
0.3752850294113159,
0.24541549384593964,
0.03588256239891052,
-0.1315302848815918,
-0.0918596163392067,
0.5690752863883972,
-0.15033242106437683,
0.1018586978316307,
... |
https://github.com/huggingface/datasets/issues/5276 | Sounds like one of the file is not a valid one, can you make sure you uploaded valid mp3 files ? | Bug in downloading common_voice data and snall chunk of it to one's own hub | ### Describe the bug
I'm trying to load the common voice dataset. Currently there is no implementation to download just par tof the data, and I need just one part of it, without downloading the entire dataset
Help please?
| Bug in downloading common_voice data and snall chunk of it to one's own hub | ### Describe the bug
I'm trying to load the common voice dataset. Currently there is no implementation to download just par tof the data, and I need just one part of it, without downloading the entire dataset
Help please?
:
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discussions/1/files
- Original:
```yaml
... | 19 | YAML integer keys are not preserved Hub server-side
After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discu... | [
-0.026737473905086517,
-0.33031290769577026,
0.08248206228017807,
0.08531218767166138,
0.39381861686706543,
-0.0846637487411499,
0.4583663046360016,
0.06925198435783386,
0.29853421449661255,
-0.023020148277282715,
-0.1234208270907402,
0.2475961297750473,
-0.23985137045383453,
0.16388651728... |
https://github.com/huggingface/datasets/issues/5275 | 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... | YAML integer keys are not preserved Hub server-side | After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discussions/1/files
- Original:
```yaml
... | 54 | YAML integer keys are not preserved Hub server-side
After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discu... | [
-0.026737473905086517,
-0.33031290769577026,
0.08248206228017807,
0.08531218767166138,
0.39381861686706543,
-0.0846637487411499,
0.4583663046360016,
0.06925198435783386,
0.29853421449661255,
-0.023020148277282715,
-0.1234208270907402,
0.2475961297750473,
-0.23985137045383453,
0.16388651728... |
https://github.com/huggingface/datasets/issues/5275 | but isn't `datasets 2.6.1` downloading files from the Hub with the corresponding tag? I thought we had something like this before | YAML integer keys are not preserved Hub server-side | After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discussions/1/files
- Original:
```yaml
... | 21 | YAML integer keys are not preserved Hub server-side
After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discu... | [
-0.026737473905086517,
-0.33031290769577026,
0.08248206228017807,
0.08531218767166138,
0.39381861686706543,
-0.0846637487411499,
0.4583663046360016,
0.06925198435783386,
0.29853421449661255,
-0.023020148277282715,
-0.1234208270907402,
0.2475961297750473,
-0.23985137045383453,
0.16388651728... |
https://github.com/huggingface/datasets/issues/5275 | 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.
But yea a year ago we had those tags, we just ended up not using them | YAML integer keys are not preserved Hub server-side | After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discussions/1/files
- Original:
```yaml
... | 42 | YAML integer keys are not preserved Hub server-side
After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discu... | [
-0.026737473905086517,
-0.33031290769577026,
0.08248206228017807,
0.08531218767166138,
0.39381861686706543,
-0.0846637487411499,
0.4583663046360016,
0.06925198435783386,
0.29853421449661255,
-0.023020148277282715,
-0.1234208270907402,
0.2475961297750473,
-0.23985137045383453,
0.16388651728... |
https://github.com/huggingface/datasets/issues/5275 | 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 | YAML integer keys are not preserved Hub server-side | After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discussions/1/files
- Original:
```yaml
... | 25 | YAML integer keys are not preserved Hub server-side
After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discu... | [
-0.026737473905086517,
-0.33031290769577026,
0.08248206228017807,
0.08531218767166138,
0.39381861686706543,
-0.0846637487411499,
0.4583663046360016,
0.06925198435783386,
0.29853421449661255,
-0.023020148277282715,
-0.1234208270907402,
0.2475961297750473,
-0.23985137045383453,
0.16388651728... |
https://github.com/huggingface/datasets/issues/5275 | So, is it OK to make the bulk edit on the Hub now or should we wait longer? If the latter, how long? | YAML integer keys are not preserved Hub server-side | After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discussions/1/files
- Original:
```yaml
... | 23 | YAML integer keys are not preserved Hub server-side
After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discu... | [
-0.026737473905086517,
-0.33031290769577026,
0.08248206228017807,
0.08531218767166138,
0.39381861686706543,
-0.0846637487411499,
0.4583663046360016,
0.06925198435783386,
0.29853421449661255,
-0.023020148277282715,
-0.1234208270907402,
0.2475961297750473,
-0.23985137045383453,
0.16388651728... |
https://github.com/huggingface/datasets/issues/5275 | 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 | YAML integer keys are not preserved Hub server-side | After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discussions/1/files
- Original:
```yaml
... | 54 | YAML integer keys are not preserved Hub server-side
After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discu... | [
-0.026737473905086517,
-0.33031290769577026,
0.08248206228017807,
0.08531218767166138,
0.39381861686706543,
-0.0846637487411499,
0.4583663046360016,
0.06925198435783386,
0.29853421449661255,
-0.023020148277282715,
-0.1234208270907402,
0.2475961297750473,
-0.23985137045383453,
0.16388651728... |
https://github.com/huggingface/datasets/issues/5275 | The bulk edit of canonical datasets (except imdb and conllpp) is running.
See e.g.: https://huggingface.co/datasets/acronym_identification/discussions/3
EDITED:
Done, except for "universal_morphologies", where I get
```
HTTPError: 413 Client Error: Payload Too Large for url: https://huggingface.co/api/valida... | YAML integer keys are not preserved Hub server-side | After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discussions/1/files
- Original:
```yaml
... | 53 | YAML integer keys are not preserved Hub server-side
After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discu... | [
-0.026737473905086517,
-0.33031290769577026,
0.08248206228017807,
0.08531218767166138,
0.39381861686706543,
-0.0846637487411499,
0.4583663046360016,
0.06925198435783386,
0.29853421449661255,
-0.023020148277282715,
-0.1234208270907402,
0.2475961297750473,
-0.23985137045383453,
0.16388651728... |
https://github.com/huggingface/datasets/issues/5275 | @lhoestq, there are 6 community datasets with YAML integer keys in their `dataset_info` `class_label`:
- indonlp/indonlu
- rcds/swiss_judgment_prediction
- Jean-Baptiste/wikiner_fr
- Bingsu/Cat_and_Dog
- taskydata/tasky_or_not
- RCC-MSU/collection3
Maybe we could open a PR on them as well? | YAML integer keys are not preserved Hub server-side | After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discussions/1/files
- Original:
```yaml
... | 36 | YAML integer keys are not preserved Hub server-side
After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discu... | [
-0.026737473905086517,
-0.33031290769577026,
0.08248206228017807,
0.08531218767166138,
0.39381861686706543,
-0.0846637487411499,
0.4583663046360016,
0.06925198435783386,
0.29853421449661255,
-0.023020148277282715,
-0.1234208270907402,
0.2475961297750473,
-0.23985137045383453,
0.16388651728... |
https://github.com/huggingface/datasets/issues/5275 | Let's do this then:
- [x] [indonlp/indonlu](https://huggingface.co/datasets/indonlp/indonlu/discussions/3)
- [x] rcds/swiss_judgment_prediction
- [x] Jean-Baptiste/wikiner_fr
- [x] Bingsu/Cat_and_Dog -> merged
- [x] taskydata/tasky_or_not (was already using quotes)
- [x] RCC-MSU/collection3
EDIT: all done :) | YAML integer keys are not preserved Hub server-side | After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discussions/1/files
- Original:
```yaml
... | 32 | YAML integer keys are not preserved Hub server-side
After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discu... | [
-0.026737473905086517,
-0.33031290769577026,
0.08248206228017807,
0.08531218767166138,
0.39381861686706543,
-0.0846637487411499,
0.4583663046360016,
0.06925198435783386,
0.29853421449661255,
-0.023020148277282715,
-0.1234208270907402,
0.2475961297750473,
-0.23985137045383453,
0.16388651728... |
https://github.com/huggingface/datasets/issues/5275 | @lhoestq I was not asking you to do it, but asking if you agree me to do it... :man_facepalming:
As I self-assigned this issue... :sweat_smile: | YAML integer keys are not preserved Hub server-side | After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discussions/1/files
- Original:
```yaml
... | 25 | YAML integer keys are not preserved Hub server-side
After an internal discussion (https://github.com/huggingface/moon-landing/issues/4563):
- YAML integer keys are not preserved server-side: they are transformed to strings
- See for example this Hub PR: https://huggingface.co/datasets/acronym_identification/discu... | [
-0.026737473905086517,
-0.33031290769577026,
0.08248206228017807,
0.08531218767166138,
0.39381861686706543,
-0.0846637487411499,
0.4583663046360016,
0.06925198435783386,
0.29853421449661255,
-0.023020148277282715,
-0.1234208270907402,
0.2475961297750473,
-0.23985137045383453,
0.16388651728... |
https://github.com/huggingface/datasets/issues/5274 | Btw, thanks very much for finding the hub rollback temporary fix and bringing the issue to our attention @KhoomeiK! | load_dataset possibly broken for gated datasets? | ### Describe the bug
When trying to download the [winoground dataset](https://huggingface.co/datasets/facebook/winoground), I get this error unless I roll back the version of huggingface-hub:
```
[/usr/local/lib/python3.7/dist-packages/huggingface_hub/utils/_validators.py](https://localhost:8080/#) in validate_rep... | 19 | load_dataset possibly broken for gated datasets?
### Describe the bug
When trying to download the [winoground dataset](https://huggingface.co/datasets/facebook/winoground), I get this error unless I roll back the version of huggingface-hub:
```
[/usr/local/lib/python3.7/dist-packages/huggingface_hub/utils/_valid... | [
-0.4157962203025818,
-0.003922760486602783,
0.032588690519332886,
0.26574575901031494,
0.26149630546569824,
-0.057288456708192825,
0.3101870119571686,
-0.02030438557267189,
0.20288017392158508,
-0.2086801379919052,
-0.4507502317428589,
-0.19844979047775269,
-0.039928555488586426,
0.4498441... |
https://github.com/huggingface/datasets/issues/5274 | 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`.
https://github.com/poloclub/diffusiondb/issues/7 | load_dataset possibly broken for gated datasets? | ### Describe the bug
When trying to download the [winoground dataset](https://huggingface.co/datasets/facebook/winoground), I get this error unless I roll back the version of huggingface-hub:
```
[/usr/local/lib/python3.7/dist-packages/huggingface_hub/utils/_validators.py](https://localhost:8080/#) in validate_rep... | 20 | load_dataset possibly broken for gated datasets?
### Describe the bug
When trying to download the [winoground dataset](https://huggingface.co/datasets/facebook/winoground), I get this error unless I roll back the version of huggingface-hub:
```
[/usr/local/lib/python3.7/dist-packages/huggingface_hub/utils/_valid... | [
-0.40456297993659973,
0.04139460623264313,
0.049647051841020584,
0.29284676909446716,
0.2792354226112366,
-0.038268111646175385,
0.4249747395515442,
-0.06315013766288757,
0.18568406999111176,
-0.16524803638458252,
-0.43654289841651917,
-0.19525548815727234,
-0.05091008543968201,
0.38398683... |
https://github.com/huggingface/datasets/issues/5274 | I fixed my issue by specifying `repo_type` in `hf_hub_url()`. https://github.com/poloclub/diffusiondb/commit/9eb91c79aaca98b0515a0ce45778b8af65b84652
I opened a PR on the Winoground's repo: https://huggingface.co/datasets/facebook/winoground/discussions/2 | load_dataset possibly broken for gated datasets? | ### Describe the bug
When trying to download the [winoground dataset](https://huggingface.co/datasets/facebook/winoground), I get this error unless I roll back the version of huggingface-hub:
```
[/usr/local/lib/python3.7/dist-packages/huggingface_hub/utils/_validators.py](https://localhost:8080/#) in validate_rep... | 19 | load_dataset possibly broken for gated datasets?
### Describe the bug
When trying to download the [winoground dataset](https://huggingface.co/datasets/facebook/winoground), I get this error unless I roll back the version of huggingface-hub:
```
[/usr/local/lib/python3.7/dist-packages/huggingface_hub/utils/_valid... | [
-0.4178113341331482,
0.027515843510627747,
0.041722021996974945,
0.2775747776031494,
0.28116151690483093,
-0.05594339221715927,
0.37349095940589905,
-0.04132677614688873,
0.20919598639011383,
-0.18582937121391296,
-0.49527508020401,
-0.22084739804267883,
-0.022110626101493835,
0.3965710699... |
https://github.com/huggingface/datasets/issues/5274 | 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. | load_dataset possibly broken for gated datasets? | ### Describe the bug
When trying to download the [winoground dataset](https://huggingface.co/datasets/facebook/winoground), I get this error unless I roll back the version of huggingface-hub:
```
[/usr/local/lib/python3.7/dist-packages/huggingface_hub/utils/_validators.py](https://localhost:8080/#) in validate_rep... | 35 | load_dataset possibly broken for gated datasets?
### Describe the bug
When trying to download the [winoground dataset](https://huggingface.co/datasets/facebook/winoground), I get this error unless I roll back the version of huggingface-hub:
```
[/usr/local/lib/python3.7/dist-packages/huggingface_hub/utils/_valid... | [
-0.43407201766967773,
0.007443740963935852,
0.030192844569683075,
0.21088933944702148,
0.2548949718475342,
-0.043477728962898254,
0.34015995264053345,
-0.003666185773909092,
0.2459324449300766,
-0.14946787059307098,
-0.3907081186771393,
-0.2331274002790451,
-0.03673965483903885,
0.42470893... |
https://github.com/huggingface/datasets/issues/5274 | huggingface_hub.utils._validators.HFValidationError: Repo id must be in the form 'repo_name' or 'namespace/repo_name':
tokenizer = AutoTokenizer.from_pretrained(ARGS.model_path, trust_remote_code=True)
Please handle automatically for local path and repo name inside, otherwise users always get confused about t... | load_dataset possibly broken for gated datasets? | ### Describe the bug
When trying to download the [winoground dataset](https://huggingface.co/datasets/facebook/winoground), I get this error unless I roll back the version of huggingface-hub:
```
[/usr/local/lib/python3.7/dist-packages/huggingface_hub/utils/_validators.py](https://localhost:8080/#) in validate_rep... | 32 | load_dataset possibly broken for gated datasets?
### Describe the bug
When trying to download the [winoground dataset](https://huggingface.co/datasets/facebook/winoground), I get this error unless I roll back the version of huggingface-hub:
```
[/usr/local/lib/python3.7/dist-packages/huggingface_hub/utils/_valid... | [
-0.4026070535182953,
0.010085448622703552,
0.0296064093708992,
0.2840164601802826,
0.2549498677253723,
-0.03881457448005676,
0.34951186180114746,
-0.05209551751613617,
0.16766712069511414,
-0.19293296337127686,
-0.4614606499671936,
-0.24034035205841064,
-0.04129769653081894,
0.394075125455... |
https://github.com/huggingface/datasets/issues/5274 | I think I'm also hitting this error, trying to load a model from a local path. | load_dataset possibly broken for gated datasets? | ### Describe the bug
When trying to download the [winoground dataset](https://huggingface.co/datasets/facebook/winoground), I get this error unless I roll back the version of huggingface-hub:
```
[/usr/local/lib/python3.7/dist-packages/huggingface_hub/utils/_validators.py](https://localhost:8080/#) in validate_rep... | 16 | load_dataset possibly broken for gated datasets?
### Describe the bug
When trying to download the [winoground dataset](https://huggingface.co/datasets/facebook/winoground), I get this error unless I roll back the version of huggingface-hub:
```
[/usr/local/lib/python3.7/dist-packages/huggingface_hub/utils/_valid... | [
-0.4335637092590332,
0.0938163697719574,
0.057825278490781784,
0.24788010120391846,
0.2946287989616394,
-0.014224976301193237,
0.43325018882751465,
-0.024157026782631874,
0.2477184683084488,
-0.20607180893421173,
-0.4578635096549988,
-0.19572114944458008,
-0.03618857264518738,
0.4589138329... |
https://github.com/huggingface/datasets/issues/5272 | Hi ! We're using the Arrow format for the datasets, and PyArrow tensors are not part of the Arrow format AFAIK:
> There is no direct support in the arrow columnar format to store Tensors as column values.
source: https://github.com/apache/arrow/issues/4802#issuecomment-508494694 | Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 40 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
0.018438346683979034,
-0.2970700263977051,
-0.13789543509483337,
0.24755112826824188,
0.25047680735588074,
0.1788550466299057,
0.6415414214134216,
0.4522418975830078,
0.16047951579093933,
0.03203722834587097,
-0.16738775372505188,
0.32685551047325134,
-0.09618178009986877,
-0.1541441082954... |
https://github.com/huggingface/datasets/issues/5272 | @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) | Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 40 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
-0.0513857938349247,
-0.27133414149284363,
-0.18894711136817932,
0.2378043234348297,
0.19903990626335144,
0.11489087343215942,
0.5828310251235962,
0.3966600000858307,
0.06404425203800201,
0.045205481350421906,
-0.056755393743515015,
0.3696766793727875,
-0.04829903692007065,
-0.113621301949... |
https://github.com/huggingface/datasets/issues/5272 | Hey @franz101 & @lhoestq!
There 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). | Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 35 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
0.057188481092453,
-0.3601831793785095,
-0.17032122611999512,
0.27301815152168274,
0.21003130078315735,
0.17189837992191315,
0.580357015132904,
0.4430302679538727,
0.11593945324420929,
0.07289421558380127,
-0.10692616552114487,
0.26417088508605957,
-0.07300233840942383,
-0.1826318055391311... |
https://github.com/huggingface/datasets/issues/5272 | 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 im... | Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 50 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
0.02933509647846222,
-0.2762486934661865,
-0.19528096914291382,
0.19531483948230743,
0.1582425832748413,
0.11916419118642807,
0.55086749792099,
0.39621639251708984,
0.05736607313156128,
0.07394518703222275,
-0.0860123485326767,
0.3299776315689087,
-0.10304255038499832,
-0.1470128297805786,... |
https://github.com/huggingface/datasets/issues/5272 | that is a great update, thank you.
it 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?
| Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 23 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
-0.0565699078142643,
-0.24875156581401825,
-0.19159096479415894,
0.1602638214826584,
0.21705842018127441,
0.17682746052742004,
0.6273614168167114,
0.38582611083984375,
0.11951598525047302,
0.07040227204561234,
-0.1241912841796875,
0.35180991888046265,
-0.08988462388515472,
-0.0561623312532... |
https://github.com/huggingface/datasets/issues/5272 | TensorArray sounds great ! Looking forward to it :)
We've had our own ExtensionArray for fixed shape tensors for a while now, hoping to see something more standardized by the arrow community.
Also super interested in the extension array for tensors of different sizes cc @mariosasko | Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 46 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
0.06654178351163864,
-0.37747490406036377,
-0.16613343358039856,
0.18838636577129364,
0.19850555062294006,
0.14027585089206696,
0.603227436542511,
0.4343041479587555,
0.179194375872612,
0.10342303663492203,
-0.05485469847917557,
0.31543266773223877,
-0.09547537565231323,
-0.153206244111061... |
https://github.com/huggingface/datasets/issues/5272 | @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.
Also, is there a tentative time frame for the variable-shape Tensor extension type? | Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 53 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
0.043579861521720886,
-0.3946131467819214,
-0.16347695887088776,
0.23242038488388062,
0.21521377563476562,
0.13309672474861145,
0.6192717552185059,
0.3833354413509369,
0.09690655767917633,
0.1788601130247116,
0.016185369342565536,
0.2568283975124359,
-0.10308155417442322,
0.043857835233211... |
https://github.com/huggingface/datasets/issues/5272 | @mariosasko please tag me in the discussion, perhaps I can contribute.
As 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? | Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 41 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
-0.0543859638273716,
-0.45368027687072754,
-0.20316684246063232,
0.23150527477264404,
0.16863687336444855,
0.07581654191017151,
0.5256624817848206,
0.3198232054710388,
0.23147261142730713,
0.1751745641231537,
-0.10954713076353073,
0.14738062024116516,
-0.12431003153324127,
0.15883940458297... |
https://github.com/huggingface/datasets/issues/5272 | pyarrow 12 is out 🎉, will have a look if I can work on it for the ExtensionArray | Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 18 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
0.037285491824150085,
-0.3342439532279968,
-0.1818915605545044,
0.24729780852794647,
0.2397068589925766,
0.18100300431251526,
0.5196181535720825,
0.4145871698856354,
0.13173264265060425,
0.10044243186712265,
-0.1117478758096695,
0.27391916513442993,
-0.05518052726984024,
-0.123753733932971... |
https://github.com/huggingface/datasets/issues/5272 | 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.
@rok We've had a couple of requests for supporting variable-shape tensors on the forum/GH, but I did not manage ... | Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 82 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
0.024113982915878296,
-0.3865979313850403,
-0.16492785513401031,
0.1982584148645401,
0.20200267434120178,
0.0881553441286087,
0.6093383431434631,
0.3891158699989319,
0.07199597358703613,
0.08206281065940857,
-0.03257215768098831,
0.3675234913825989,
-0.08546142280101776,
-0.148334383964538... |
https://github.com/huggingface/datasets/issues/5272 | > @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://... | Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 79 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
-0.008001565933227539,
-0.4088495373725891,
-0.15685778856277466,
0.180075541138649,
0.2072976529598236,
0.0651898980140686,
0.5626472234725952,
0.4157724976539612,
0.029195889830589294,
0.1282946616411209,
-0.12156055867671967,
0.3279113173484802,
-0.10012753307819366,
-0.0772008448839187... |
https://github.com/huggingface/datasets/issues/5272 | 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 ! | Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 34 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
-0.06251892447471619,
-0.30171531438827515,
-0.15307942032814026,
0.21659913659095764,
0.22312955558300018,
0.1660510003566742,
0.6269474625587463,
0.41092318296432495,
0.10571213811635971,
0.021507494151592255,
-0.15369674563407898,
0.3499835729598999,
-0.10253965109586716,
-0.09003025293... |
https://github.com/huggingface/datasets/issues/5272 | 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. | Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 25 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
0.011571317911148071,
-0.30237042903900146,
-0.1396525502204895,
0.25233471393585205,
0.25077107548713684,
0.15381303429603577,
0.5770966410636902,
0.42510122060775757,
0.09155572950839996,
0.05140848457813263,
-0.06717948615550995,
0.31220871210098267,
-0.04894430190324783,
-0.16544932126... |
https://github.com/huggingface/datasets/issues/5272 | 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. | Use pyarrow Tensor dtype | ### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.from_numpy(x, dim_names=["dim1... | 26 | Use pyarrow Tensor dtype
### Feature request
I was going the discussion of converting tensors to lists.
Is there a way to leverage pyarrow's Tensors for nested arrays / embeddings?
For example:
```python
import pyarrow as pa
import numpy as np
x = np.array([[2, 2, 4], [4, 5, 100]], np.int32)
pa.Tensor.fro... | [
0.01900598406791687,
-0.28309503197669983,
-0.1886107325553894,
0.22200961410999298,
0.2580498456954956,
0.13851699233055115,
0.5342298150062561,
0.38313522934913635,
0.14036065340042114,
0.08998771011829376,
-0.08856257796287537,
0.3163338005542755,
-0.07234177738428116,
-0.14722293615341... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.