Datasets:

asoria

/

dataset-viewer-github-issues

like 0

Fix dataset name when decreasing metrics

Fix dataset name when decreasing metrics:

closed

2024-07-04T19:58:29Z

2024-07-08T12:04:37Z

2024-07-04T22:40:16Z

AndreaFrancis

[Modalities] Account for image URLs dataset for Image modality

right now datasets like https://huggingface.co/datasets/CaptionEmporium/coyo-hd-11m-llavanext are missing the Image modality even though they have image URLs

[Modalities] Account for image URLs dataset for Image modality: right now datasets like https://huggingface.co/datasets/CaptionEmporium/coyo-hd-11m-llavanext are missing the Image modality even though they have image URLs

closed

2024-07-04T16:28:28Z

2024-07-15T16:48:11Z

2024-07-15T16:48:10Z

lhoestq

Add threshold to modalities from filetypes

Fix modalities false positives for - https://huggingface.co/datasets/chenxx1/jia - https://huggingface.co/datasets/proj-persona/PersonaHub - https://huggingface.co/datasets/BAAI/Infinity-Instruct - https://huggingface.co/datasets/m-a-p/COIG-CQIA - https://huggingface.co/datasets/Magpie-Align/Magpie-Pro-300K-Filtered I added two thresholds to get less false positives: - one general threshold to ignore file types that are <10% of the files - one additional threshold specific to images in presence of csv/json/parquet, since images are often used in README as figures cc @severo This should take care of most false positives, and we can refine later if needed

Add threshold to modalities from filetypes: Fix modalities false positives for - https://huggingface.co/datasets/chenxx1/jia - https://huggingface.co/datasets/proj-persona/PersonaHub - https://huggingface.co/datasets/BAAI/Infinity-Instruct - https://huggingface.co/datasets/m-a-p/COIG-CQIA - https://huggingface.co/datasets/Magpie-Align/Magpie-Pro-300K-Filtered I added two thresholds to get less false positives: - one general threshold to ignore file types that are <10% of the files - one additional threshold specific to images in presence of csv/json/parquet, since images are often used in README as figures cc @severo This should take care of most false positives, and we can refine later if needed

closed

2024-07-04T13:23:59Z

2024-07-04T15:19:47Z

2024-07-04T15:19:45Z

lhoestq

WIP: Try to get languages from librarian bot PR for FTS

This PR is still in progress, but it is a suggestion about how to get the language from open PRs from the librarian bot. Please let me know what you think. Pending: tests, refactor. For https://huggingface.co/datasets/Osumansan/data-poison/commit/22432ba97e6c559891bd82ca084496a7f8a6699f.diff , it was able to identify 'ko' language, but since it is not supported by DuckDB, it assigns 'none' as stemmer. cc. @davanstrien

WIP: Try to get languages from librarian bot PR for FTS : This PR is still in progress, but it is a suggestion about how to get the language from open PRs from the librarian bot. Please let me know what you think. Pending: tests, refactor. For https://huggingface.co/datasets/Osumansan/data-poison/commit/22432ba97e6c559891bd82ca084496a7f8a6699f.diff , it was able to identify 'ko' language, but since it is not supported by DuckDB, it assigns 'none' as stemmer. cc. @davanstrien

closed

2024-07-03T17:05:20Z

2024-07-05T12:49:45Z

2024-07-05T12:49:45Z

AndreaFrancis

Add duration to cached steps

Will close https://github.com/huggingface/dataset-viewer/issues/2892 As suggested in https://github.com/huggingface/dataset-viewer/pull/2908#pullrequestreview-2126425919, this adds duration field to cached responses. Duration is computed using `started_at` field of a corresponding job. Note that this PR adds new fields to major libcommon dtos `JobInfo` and `JobResult`.

Add duration to cached steps: Will close https://github.com/huggingface/dataset-viewer/issues/2892 As suggested in https://github.com/huggingface/dataset-viewer/pull/2908#pullrequestreview-2126425919, this adds duration field to cached responses. Duration is computed using `started_at` field of a corresponding job. Note that this PR adds new fields to major libcommon dtos `JobInfo` and `JobResult`.

closed

2024-07-03T14:17:47Z

2024-07-09T13:06:37Z

2024-07-09T13:06:35Z

polinaeterna

Use placeholder revision in urls in cached responses

Previously, we were using the dataset revision in the URLs of image/audio files of cached responses of /first-rows. However when a dataset gets its README updated, we update the `dataset_git_revision` of the cache entries and the location of the image/audio files on S3 but we don't modify the revision in the URLs in the cached response. This resulted in the Viewer not showing the images after modifying the readme of a dataset. <img width="1020" alt="image" src="https://github.com/huggingface/dataset-viewer/assets/42851186/6242d9d6-bc2e-42b1-9edd-9f01173b6cd8"> I fixed that for future datasets by not using the revision in the URLs anymore and use a placeholder that is replaced by `dataset_git_revision` when the cached response is accessed ## Implementation details I modified the URL Signer logic to also insert the revision in the URL and renamed it to a URL Preparator. It takes care of inserting the revision and signing the URLs. close https://github.com/huggingface/dataset-viewer/issues/2965

Use placeholder revision in urls in cached responses: Previously, we were using the dataset revision in the URLs of image/audio files of cached responses of /first-rows. However when a dataset gets its README updated, we update the `dataset_git_revision` of the cache entries and the location of the image/audio files on S3 but we don't modify the revision in the URLs in the cached response. This resulted in the Viewer not showing the images after modifying the readme of a dataset. <img width="1020" alt="image" src="https://github.com/huggingface/dataset-viewer/assets/42851186/6242d9d6-bc2e-42b1-9edd-9f01173b6cd8"> I fixed that for future datasets by not using the revision in the URLs anymore and use a placeholder that is replaced by `dataset_git_revision` when the cached response is accessed ## Implementation details I modified the URL Signer logic to also insert the revision in the URL and renamed it to a URL Preparator. It takes care of inserting the revision and signing the URLs. close https://github.com/huggingface/dataset-viewer/issues/2965

closed

2024-07-03T12:33:20Z

2024-07-15T17:27:48Z

2024-07-15T17:27:46Z

lhoestq

Viewer doesn't show images properly after a smart update

we move the images on s3 in case of modified readme, but we don't update the location of the images in the first-rows responses

Viewer doesn't show images properly after a smart update: we move the images on s3 in case of modified readme, but we don't update the location of the images in the first-rows responses

closed

2024-07-03T12:33:07Z

2024-07-15T17:27:47Z

2024-07-15T17:27:47Z

lhoestq

Viewer shows outdated cache after renaming a repo and creating a new one with the old name

Reported by @lewtun (internal link: https://huggingface.slack.com/archives/C02EMARJ65P/p1719818961944059): > If I rename a dataset via the UI from D -> D' and then create a new dataset with the same name D, I seem to get a copy instead of an empty dataset > Indeed it was the dataset viewer showing a cached result - the git history is clean and there's no files in the new dataset repo

Viewer shows outdated cache after renaming a repo and creating a new one with the old name: Reported by @lewtun (internal link: https://huggingface.slack.com/archives/C02EMARJ65P/p1719818961944059): > If I rename a dataset via the UI from D -> D' and then create a new dataset with the same name D, I seem to get a copy instead of an empty dataset > Indeed it was the dataset viewer showing a cached result - the git history is clean and there's no files in the new dataset repo

open

2024-07-02T07:05:53Z

2024-08-23T14:16:39Z

albertvillanova

Fix ISO 639-1 mapping for stemming

Fix ISO 639-1 mapping for stemming:

closed

2024-07-01T15:08:12Z

2024-07-01T15:33:48Z

2024-07-01T15:33:46Z

AndreaFrancis

Removing has_fts field from split-duckdb-index

Context: https://github.com/huggingface/dataset-viewer/pull/2928#discussion_r1652733919 I'm removing the `has_fts` field from `split-duckdb-index`, given that now, the `stemmer` field will indicate if the split supports the feature. `stemmer`=None means no FTS; otherwise, it does.

Removing has_fts field from split-duckdb-index: Context: https://github.com/huggingface/dataset-viewer/pull/2928#discussion_r1652733919 I'm removing the `has_fts` field from `split-duckdb-index`, given that now, the `stemmer` field will indicate if the split supports the feature. `stemmer`=None means no FTS; otherwise, it does.

closed

2024-06-27T16:08:56Z

2024-07-01T15:58:18Z

2024-07-01T15:58:17Z

AndreaFrancis

update test_plan_job_creation_and_termination

...to fix the CI after https://github.com/huggingface/dataset-viewer/pull/2958

update test_plan_job_creation_and_termination: ...to fix the CI after https://github.com/huggingface/dataset-viewer/pull/2958

closed

2024-06-27T15:11:13Z

2024-06-28T08:17:24Z

2024-06-27T15:26:52Z

lhoestq

Detect rename in smart update

reported in https://huggingface.co/datasets/crossingminds/shopping-queries-image-dataset/discussions/2

Detect rename in smart update: reported in https://huggingface.co/datasets/crossingminds/shopping-queries-image-dataset/discussions/2

closed

2024-06-27T14:50:49Z

2024-06-27T15:38:41Z

2024-06-27T15:38:39Z

lhoestq

add diagram to docs

fixes #2956

add diagram to docs: fixes #2956

closed

2024-06-27T13:09:52Z

2024-06-27T15:22:21Z

2024-06-27T15:22:18Z

severo

Remove blocked only job types

fixes #2957

Remove blocked only job types: fixes #2957

closed

2024-06-27T11:59:46Z

2024-06-27T13:17:06Z

2024-06-27T13:17:04Z

severo

Remove logic for `WORKER_JOB_TYPES_BLOCKED` and `WORKER_JOB_TYPES_ONLY`

they are not used anymore (empty lists).

Remove logic for `WORKER_JOB_TYPES_BLOCKED` and `WORKER_JOB_TYPES_ONLY`: they are not used anymore (empty lists).

closed

2024-06-27T11:47:27Z

2024-06-27T13:17:05Z

2024-06-27T13:17:05Z

severo

Elaborate a diagram that describes the queues/prioritization logic

it would be useful to discuss issues like https://github.com/huggingface/dataset-viewer/issues/2955

Elaborate a diagram that describes the queues/prioritization logic: it would be useful to discuss issues like https://github.com/huggingface/dataset-viewer/issues/2955

closed

2024-06-27T11:11:22Z

2024-06-27T15:22:19Z

2024-06-27T15:22:19Z

severo

prioritize jobs from trendy/important datasets

internal Slack discussion: https://huggingface.slack.com/archives/C04L6P8KNQ5/p1719482620323589?thread_ts=1719418419.785649&cid=C04L6P8KNQ5 > prioritization of datasets should probably be based on some popularity signal like number of likes or traffic to the dataset page in the future

prioritize jobs from trendy/important datasets: internal Slack discussion: https://huggingface.slack.com/archives/C04L6P8KNQ5/p1719482620323589?thread_ts=1719418419.785649&cid=C04L6P8KNQ5 > prioritization of datasets should probably be based on some popularity signal like number of likes or traffic to the dataset page in the future

open

2024-06-27T11:09:55Z

2024-07-30T16:28:58Z

severo

Smart update on all datasets

I did a bunch of test on https://huggingface.co/datasets/datasets-maintainers/test-smart-update and it works fine: edits to README.md unrelated to the `configs` don't trigger a full recomputation of the viewer :) I also tested with image data and they are correctly handled.

Smart update on all datasets: I did a bunch of test on https://huggingface.co/datasets/datasets-maintainers/test-smart-update and it works fine: edits to README.md unrelated to the `configs` don't trigger a full recomputation of the viewer :) I also tested with image data and they are correctly handled.

closed

2024-06-26T16:57:18Z

2024-06-26T17:00:02Z

2024-06-26T17:00:00Z

lhoestq

add /admin/blocked-datasets endpoint

I will use it in https://observablehq.com/@huggingface/datasets-server-jobs-queue to be able to understand better the current jobs.

add /admin/blocked-datasets endpoint: I will use it in https://observablehq.com/@huggingface/datasets-server-jobs-queue to be able to understand better the current jobs.

closed

2024-06-26T16:46:35Z

2024-06-26T19:20:12Z

2024-06-26T19:20:10Z

severo

No cache in smart update

...since it was causing ``` smart_update_dataset failed with PermissionError: [Errno 13] Permission denied: '/.cache' ``` I used the HfFileSystem instead to read the README.md files, since it doesn't use caching

No cache in smart update: ...since it was causing ``` smart_update_dataset failed with PermissionError: [Errno 13] Permission denied: '/.cache' ``` I used the HfFileSystem instead to read the README.md files, since it doesn't use caching

closed

2024-06-26T14:43:27Z

2024-06-26T16:05:23Z

2024-06-26T16:05:21Z

lhoestq

Add estimated_num_rows in openapi

Add estimated_num_rows in openapi:

closed

2024-06-25T16:46:42Z

2024-07-25T12:02:34Z

2024-07-25T12:02:33Z

lhoestq

add missing migration for estimated_num_rows

add missing migration for estimated_num_rows:

closed

2024-06-25T16:25:31Z

2024-06-25T16:28:44Z

2024-06-25T16:28:44Z

lhoestq

Ignore blocked datasets in WorkerSize metrics for auto scaling

Fix for https://github.com/huggingface/dataset-viewer/issues/2945

Ignore blocked datasets in WorkerSize metrics for auto scaling: Fix for https://github.com/huggingface/dataset-viewer/issues/2945

closed

2024-06-25T15:26:26Z

2024-06-26T16:23:57Z

2024-06-26T15:15:52Z

AndreaFrancis

Exclude blocked datasets from Job metrics

Fix for https://github.com/huggingface/dataset-viewer/issues/2945

Exclude blocked datasets from Job metrics: Fix for https://github.com/huggingface/dataset-viewer/issues/2945

closed

2024-06-25T00:38:56Z

2024-06-25T15:11:57Z

2024-06-25T15:11:56Z

AndreaFrancis

update indexes

I think the old ones will remain. I'll remove them manually... These two indexes have been proposed by mongo cloud. The reason is: https://github.com/huggingface/dataset-viewer/pull/2933/files#diff-4c951d0a5e21ef5c719bc392169f41e726461028dfd8e049778fedff37ba38c8R422

update indexes: I think the old ones will remain. I'll remove them manually... These two indexes have been proposed by mongo cloud. The reason is: https://github.com/huggingface/dataset-viewer/pull/2933/files#diff-4c951d0a5e21ef5c719bc392169f41e726461028dfd8e049778fedff37ba38c8R422

closed

2024-06-24T22:08:29Z

2024-06-24T22:11:13Z

2024-06-24T22:11:12Z

severo

add cudf to toctree

Follow up to https://github.com/huggingface/dataset-viewer/pull/2941 I realized the docs were not building in the CI: https://github.com/huggingface/dataset-viewer/actions/runs/9648374360/job/26609396615 Apologies for not checking this in my prior PR. 

add cudf to toctree: Follow up to https://github.com/huggingface/dataset-viewer/pull/2941 I realized the docs were not building in the CI: https://github.com/huggingface/dataset-viewer/actions/runs/9648374360/job/26609396615 Apologies for not checking this in my prior PR. 

closed

2024-06-24T16:22:08Z

2024-06-24T18:33:22Z

2024-06-24T18:30:42Z

raybellwaves

Add "blocked/not blocked" in job count metrics

Now that we block datasets, the job count metrics are a bit misleading, because they still include the jobs of blocked datasets. We need to be able to filter them out, because they are outside of the queue during the blockage.

Add "blocked/not blocked" in job count metrics: Now that we block datasets, the job count metrics are a bit misleading, because they still include the jobs of blocked datasets. We need to be able to filter them out, because they are outside of the queue during the blockage.

open

2024-06-24T15:44:48Z

2024-07-30T15:48:54Z

severo

Remove old get_df code

Remove old get_df code:

closed

2024-06-24T13:27:36Z

2024-06-24T14:38:04Z

2024-06-24T14:38:03Z

AndreaFrancis

Support `Sequence()` features in Croissant crumbs.

WIP: still checking that it works as intended on mlcroissant side.

Support `Sequence()` features in Croissant crumbs.: WIP: still checking that it works as intended on mlcroissant side.

closed

2024-06-24T13:08:51Z

2024-07-22T11:01:13Z

2024-07-22T11:00:38Z

marcenacp

Increase blockage duration

blocked for 1h -> 6h based on the last 6h -> 1h The effect should be to block more datasets, more quickly, and for a longer time. Currently, all the resources are still dedicated to datasets that are updated every xxx minutes.

Increase blockage duration: blocked for 1h -> 6h based on the last 6h -> 1h The effect should be to block more datasets, more quickly, and for a longer time. Currently, all the resources are still dedicated to datasets that are updated every xxx minutes.

closed

2024-06-24T10:20:46Z

2024-06-24T10:46:13Z

2024-06-24T10:46:12Z

severo

add cudf example

Firstly, thanks a lot for this project and hosting so many datasets! This PR add an example for how to read in data using cudf. This can be useful if you have access to a GPU and want to use the GPU to accelerate any ETL. The code works and can be testing in this google colab notebook: https://colab.research.google.com/drive/1KrhARlJMPoaIpHqE9X2fMQFcYTLffrem?usp=sharing

add cudf example: Firstly, thanks a lot for this project and hosting so many datasets! This PR add an example for how to read in data using cudf. This can be useful if you have access to a GPU and want to use the GPU to accelerate any ETL. The code works and can be testing in this google colab notebook: https://colab.research.google.com/drive/1KrhARlJMPoaIpHqE9X2fMQFcYTLffrem?usp=sharing

closed

2024-06-24T01:54:48Z

2024-06-24T15:45:57Z

2024-06-24T15:45:57Z

raybellwaves

Add num_rows estimate in hub_cache

Added estimated_num_rows to config-size, dataset-size, and I also updated `hub_cache` to use estimated_num_rows if possible as `num_rows` (this way no need to modify anything in moon) TODO - [x] mongodb migration of size jobs - [x] update tests - [x] support mix of partial and exact num_rows in config-size and dataset-size - [x] revert sylvain's suggestion to add "num_rows_source" that is actually not needed

Add num_rows estimate in hub_cache: Added estimated_num_rows to config-size, dataset-size, and I also updated `hub_cache` to use estimated_num_rows if possible as `num_rows` (this way no need to modify anything in moon) TODO - [x] mongodb migration of size jobs - [x] update tests - [x] support mix of partial and exact num_rows in config-size and dataset-size - [x] revert sylvain's suggestion to add "num_rows_source" that is actually not needed

closed

2024-06-21T15:41:31Z

2024-06-25T16:18:28Z

2024-06-25T16:12:54Z

lhoestq

fix flaky test gen_kwargs

fix flaky test gen_kwargs:

closed

2024-06-21T15:05:48Z

2024-06-21T15:08:18Z

2024-06-21T15:08:17Z

lhoestq

Do not keep DataFrames in memory in orchestrator classes

Do not keep unnecessary DataFrames in memory in orchestrator classes and instead forward them for use only in class instantiation or reset them after being forwarded. This PR is related to: - #2921

Do not keep DataFrames in memory in orchestrator classes: Do not keep unnecessary DataFrames in memory in orchestrator classes and instead forward them for use only in class instantiation or reset them after being forwarded. This PR is related to: - #2921

closed

2024-06-21T14:18:17Z

2024-07-29T15:05:03Z

2024-07-29T15:05:02Z

albertvillanova

Enable estimate info (size) on all datasets

TODO before merging: - [x] merge https://github.com/huggingface/dataset-viewer/pull/2932 - [x] test json in prod (allenai/c4 en.noblocklist: correct relative error of only 0.02%) - [x] test webdataset in prod (datasets-maintainers/small-publaynet-wds-10x: correct with relative error of only 0.07%)

Enable estimate info (size) on all datasets: TODO before merging: - [x] merge https://github.com/huggingface/dataset-viewer/pull/2932 - [x] test json in prod (allenai/c4 en.noblocklist: correct relative error of only 0.02%) - [x] test webdataset in prod (datasets-maintainers/small-publaynet-wds-10x: correct with relative error of only 0.07%)

closed

2024-06-21T10:33:27Z

2024-06-24T13:14:51Z

2024-06-21T13:25:22Z

lhoestq

Update urllib3 to 1.26.19 and 2.2.2 to fix vulnerability

Update urllib3 to 1.26.19 and 2.2.2 to fix vulnerability. This PR will close 14 Dependabot alerts.

Update urllib3 to 1.26.19 and 2.2.2 to fix vulnerability: Update urllib3 to 1.26.19 and 2.2.2 to fix vulnerability. This PR will close 14 Dependabot alerts.

closed

2024-06-21T09:14:26Z

2024-06-25T11:50:49Z

2024-06-25T11:50:49Z

albertvillanova

divide the rate-limit budget by 5

divide the rate-limit budget by 5:

closed

2024-06-21T07:46:31Z

2024-06-21T07:51:35Z

2024-06-21T07:51:34Z

severo

Update scikit-learn to 1.5.0 to fix vulnerability

Update scikit-learn to 1.5.0 to fix vulnerability. This will close 12 Dependabot alerts.

Update scikit-learn to 1.5.0 to fix vulnerability: Update scikit-learn to 1.5.0 to fix vulnerability. This will close 12 Dependabot alerts.

closed

2024-06-21T07:31:32Z

2024-06-21T08:21:29Z

2024-06-21T08:21:28Z

albertvillanova

create datasetBlockages collection + block datasets

We apply rate limiting on the jobs, based on the total duration in a window (see https://github.com/huggingface/dataset-viewer/issues/2279#issuecomment-2178655627). Follows #2931

create datasetBlockages collection + block datasets: We apply rate limiting on the jobs, based on the total duration in a window (see https://github.com/huggingface/dataset-viewer/issues/2279#issuecomment-2178655627). Follows #2931

closed

2024-06-20T16:12:52Z

2024-06-20T20:48:38Z

2024-06-20T20:48:37Z

severo

Fix estimate info for zip datasets

I simply had to track metadata reads only once to fix the estimator. (otherwise every file opened in a zip archive triggers an additional read of the metadata with the central directory of the zip file that prevents the estimator from converging) ex: locally and on only 100MB of parquet conversion (prod is 5GB), it estimates 47871 examples on https://huggingface.co/datasets/datasets-maintainers/test_many_zip_with_images_MSCOCO_zip (true value is exactly 50k) Unrelared, but I noticed that parquet conversion of zip files is super slooooooowww, we'll have to improve that at one point because it can surely take more than 40min to run. When tested in prod, an error happens before that though (JobManagerCrashedError) which suggests a termination by kubernetes.

Fix estimate info for zip datasets: I simply had to track metadata reads only once to fix the estimator. (otherwise every file opened in a zip archive triggers an additional read of the metadata with the central directory of the zip file that prevents the estimator from converging) ex: locally and on only 100MB of parquet conversion (prod is 5GB), it estimates 47871 examples on https://huggingface.co/datasets/datasets-maintainers/test_many_zip_with_images_MSCOCO_zip (true value is exactly 50k) Unrelared, but I noticed that parquet conversion of zip files is super slooooooowww, we'll have to improve that at one point because it can surely take more than 40min to run. When tested in prod, an error happens before that though (JobManagerCrashedError) which suggests a termination by kubernetes.

closed

2024-06-20T15:00:39Z

2024-06-21T13:13:05Z

2024-06-21T13:13:05Z

lhoestq

Create pastJobs collection

It will be used to apply rate limiting on the jobs, based on the total duration in a window (see https://github.com/huggingface/dataset-viewer/issues/2279#issuecomment-2178655627).

Create pastJobs collection: It will be used to apply rate limiting on the jobs, based on the total duration in a window (see https://github.com/huggingface/dataset-viewer/issues/2279#issuecomment-2178655627).

closed

2024-06-20T13:41:00Z

2024-06-20T20:24:49Z

2024-06-20T20:24:48Z

severo

[refactoring] split queue.py in 3 modules

To prepare https://github.com/huggingface/dataset-viewer/issues/2279

[refactoring] split queue.py in 3 modules: To prepare https://github.com/huggingface/dataset-viewer/issues/2279

closed

2024-06-20T12:11:09Z

2024-06-20T12:39:17Z

2024-06-20T12:39:16Z

severo

Use current priority for children jobs

When we change the priority of a job manually after it has started, we want the children jobs to use the same priority

Use current priority for children jobs: When we change the priority of a job manually after it has started, we want the children jobs to use the same priority

closed

2024-06-20T09:07:05Z

2024-06-20T09:18:09Z

2024-06-20T09:18:08Z

severo

FTS: Add specific stemmer for monolingual datasets

FTS: Add specific stemmer for monolingual datasets:

closed

2024-06-19T21:32:29Z

2024-06-26T14:14:54Z

2024-06-26T14:14:52Z

AndreaFrancis

Separate expected errors from unexpected ones in Grafana

We should never have: `UnexpectedError`, `PreviousStepFormatError` Also: some error should be temporary: `PreviousStepNotReady`, etc. Instead of one chart with the three kinds of errors, we should have: - one with the expected errors - one with the transitory errors - one with the unexpected errors By the way, I think we could create intermediary classes (`ExpectedError`, `UnexpectedError`, `TransitoryError`), so that it's clear which error pertains to which category

Separate expected errors from unexpected ones in Grafana: We should never have: `UnexpectedError`, `PreviousStepFormatError` Also: some error should be temporary: `PreviousStepNotReady`, etc. Instead of one chart with the three kinds of errors, we should have: - one with the expected errors - one with the transitory errors - one with the unexpected errors By the way, I think we could create intermediary classes (`ExpectedError`, `UnexpectedError`, `TransitoryError`), so that it's clear which error pertains to which category

open

2024-06-19T20:40:19Z

2024-06-19T20:40:28Z

severo

only raise error in config-is-valid if format is bad

Should remove the remaining PreviousStepFormatError: https://github.com/huggingface/dataset-viewer/issues/2433#issuecomment-2179014134

only raise error in config-is-valid if format is bad: Should remove the remaining PreviousStepFormatError: https://github.com/huggingface/dataset-viewer/issues/2433#issuecomment-2179014134

closed

2024-06-19T20:27:44Z

2024-06-20T09:17:53Z

2024-06-20T09:17:53Z

severo

Reorder and hide columns within dataset viewer

# Problem When doing some basic vibe checks for datasets, I realized that the order in which columns are shown might not always be useful for viewing and exploring the data. I might want to quickly show `chosen_model` next to `chosen_response` and `chosen_avg_rating` and continue exploring based on that. Also, this representation might different based on personal preference and search session. # Related I noticed an earlier discussion here https://github.com/huggingface/dataset-viewer/issues/2639 (thanks @davanstrien) but it does not fully align with that ideation. # Implementation I see it more as an on-the-fly UI thing than a configurable Yaml thing (proposed in #2639), similar to re-ordering a tab in a browser. Based on this, I can also see an option to potentially hide useless columns. # Example Based on [UltraFeedback](https://huggingface.co/datasets/argilla/ultrafeedback-binarized-preferences/viewer/default/train) <img width="1465" alt="image" src="https://github.com/huggingface/dataset-viewer/assets/25269220/aac80172-969f-45b6-8b9a-513ace626320">

Reorder and hide columns within dataset viewer : # Problem When doing some basic vibe checks for datasets, I realized that the order in which columns are shown might not always be useful for viewing and exploring the data. I might want to quickly show `chosen_model` next to `chosen_response` and `chosen_avg_rating` and continue exploring based on that. Also, this representation might different based on personal preference and search session. # Related I noticed an earlier discussion here https://github.com/huggingface/dataset-viewer/issues/2639 (thanks @davanstrien) but it does not fully align with that ideation. # Implementation I see it more as an on-the-fly UI thing than a configurable Yaml thing (proposed in #2639), similar to re-ordering a tab in a browser. Based on this, I can also see an option to potentially hide useless columns. # Example Based on [UltraFeedback](https://huggingface.co/datasets/argilla/ultrafeedback-binarized-preferences/viewer/default/train) <img width="1465" alt="image" src="https://github.com/huggingface/dataset-viewer/assets/25269220/aac80172-969f-45b6-8b9a-513ace626320">

open

2024-06-19T17:32:33Z

2024-06-19T20:43:39Z

davidberenstein1957

Delete canonical datasets

The cache still contains entries for canonical datasets that have been moved to their own namespace (see https://github.com/huggingface/dataset-viewer/issues/2478#issuecomment-2179018465 for a list). We must delete the cache entries (+ assets/cached assets/jobs)

Delete canonical datasets: The cache still contains entries for canonical datasets that have been moved to their own namespace (see https://github.com/huggingface/dataset-viewer/issues/2478#issuecomment-2179018465 for a list). We must delete the cache entries (+ assets/cached assets/jobs)

closed

2024-06-19T15:51:11Z

2024-07-30T16:29:59Z

2024-07-30T16:29:59Z

severo

Fix estimate info allow_list

Fix estimate info allow_list:

closed

2024-06-19T15:29:58Z

2024-06-19T15:30:07Z

2024-06-19T15:30:06Z

lhoestq

admin-ui: Do not mark gated datasets as error

In the admin-ui, some datasets have all the features working (is-valid) but are shown as errors. This is due to gated datasets; I just added another sign to identify those and let us know that the issue is unrelated to jobs not being correctly processed. Another question is if we should consider this dataset as part of the coverage metric since we need to know the value. Or maybe there is another way to automatically subscribe to the admin-UI user token for those datasets?

admin-ui: Do not mark gated datasets as error: In the admin-ui, some datasets have all the features working (is-valid) but are shown as errors. This is due to gated datasets; I just added another sign to identify those and let us know that the issue is unrelated to jobs not being correctly processed. Another question is if we should consider this dataset as part of the coverage metric since we need to know the value. Or maybe there is another way to automatically subscribe to the admin-UI user token for those datasets?

closed

2024-06-18T22:48:21Z

2024-06-19T14:43:09Z

2024-06-19T14:43:08Z

AndreaFrancis

Do not keep DataFrames in memory in State classes

Do not keep unnecessary DataFrames in memory in State classes and instead forward them for use only in class instantiation. This PR reduces memory use by avoiding keeping unnecessary DataFrames in all State classes. This PR supersedes (neither copies nor views are longer necessary): - #2903

Do not keep DataFrames in memory in State classes: Do not keep unnecessary DataFrames in memory in State classes and instead forward them for use only in class instantiation. This PR reduces memory use by avoiding keeping unnecessary DataFrames in all State classes. This PR supersedes (neither copies nor views are longer necessary): - #2903

closed

2024-06-18T16:27:41Z

2024-06-19T05:46:09Z

2024-06-19T05:46:08Z

albertvillanova

order the steps alphabetically

fixes #2917

order the steps alphabetically: fixes #2917

closed

2024-06-18T13:48:37Z

2024-06-18T13:48:53Z

2024-06-18T13:48:52Z

severo

Do not propagate error for is valid and hub cache

We always want to have a "status", even if some of the previous steps are not available.

Do not propagate error for is valid and hub cache: We always want to have a "status", even if some of the previous steps are not available.

closed

2024-06-18T13:44:32Z

2024-06-19T15:09:01Z

2024-06-19T15:09:00Z

severo

The "dataset-hub-cache" and "dataset-is-valid" steps should always return a value

For example, we detect that the dataset `nkp37/OpenVid-1M` has Image and Video modalities (steps `dataset-filetypes` and `dataset-modalities`), but because the datasets library fails to list the configs, the following steps also return an error.

The "dataset-hub-cache" and "dataset-is-valid" steps should always return a value: For example, we detect that the dataset `nkp37/OpenVid-1M` has Image and Video modalities (steps `dataset-filetypes` and `dataset-modalities`), but because the datasets library fails to list the configs, the following steps also return an error.

closed

2024-06-18T11:13:41Z

2024-06-19T15:09:01Z

2024-06-19T15:09:01Z

severo

admin UI: automatically fill the steps list

Currently, the step `dataset-filetypes` is absent from the job types list <img width="521" alt="Capture d’écran 2024-06-18 à 13 09 01" src="https://github.com/huggingface/dataset-viewer/assets/1676121/2fc31de2-63d9-46d2-81f8-cc79dc78e868"> The list should be created automatically from the processing graph + sorted alphabetically

admin UI: automatically fill the steps list: Currently, the step `dataset-filetypes` is absent from the job types list <img width="521" alt="Capture d’écran 2024-06-18 à 13 09 01" src="https://github.com/huggingface/dataset-viewer/assets/1676121/2fc31de2-63d9-46d2-81f8-cc79dc78e868"> The list should be created automatically from the processing graph + sorted alphabetically

closed

2024-06-18T11:09:31Z

2024-06-18T13:49:31Z

2024-06-18T13:48:53Z

severo

[modality detection] One image in the repo -> Image modality

See https://huggingface.co/datasets/BAAI/Infinity-Instruct/tree/main <img width="465" alt="Capture d’écran 2024-06-18 à 12 55 35" src="https://github.com/huggingface/dataset-viewer/assets/1676121/c00b63a2-f1bd-4957-b56e-425c0a99f149"> <img width="469" alt="Capture d’écran 2024-06-18 à 12 55 49" src="https://github.com/huggingface/dataset-viewer/assets/1676121/4397834e-893f-4c28-b212-948504430c40"> It contains 5 diagram images unrelated to the dataset itself, which is a JSONL file. But the presence of these images triggers the Image modality. <img width="614" alt="Capture d’écran 2024-06-18 à 12 56 55" src="https://github.com/huggingface/dataset-viewer/assets/1676121/312181d9-334f-45af-a4e9-c5dc9371fd64">

[modality detection] One image in the repo -> Image modality: See https://huggingface.co/datasets/BAAI/Infinity-Instruct/tree/main <img width="465" alt="Capture d’écran 2024-06-18 à 12 55 35" src="https://github.com/huggingface/dataset-viewer/assets/1676121/c00b63a2-f1bd-4957-b56e-425c0a99f149"> <img width="469" alt="Capture d’écran 2024-06-18 à 12 55 49" src="https://github.com/huggingface/dataset-viewer/assets/1676121/4397834e-893f-4c28-b212-948504430c40"> It contains 5 diagram images unrelated to the dataset itself, which is a JSONL file. But the presence of these images triggers the Image modality. <img width="614" alt="Capture d’écran 2024-06-18 à 12 56 55" src="https://github.com/huggingface/dataset-viewer/assets/1676121/312181d9-334f-45af-a4e9-c5dc9371fd64">

open

2024-06-18T10:57:14Z

2024-06-18T10:57:21Z

severo

Bump urllib3 from 2.0.7 to 2.2.2 in /docs in the pip group across 1 directory

Bumps the pip group with 1 update in the /docs directory: [urllib3](https://github.com/urllib3/urllib3). Updates `urllib3` from 2.0.7 to 2.2.2 <details> <summary>Release notes</summary> <p><em>Sourced from <a href="https://github.com/urllib3/urllib3/releases">urllib3's releases</a>.</em></p> <blockquote> <h2>2.2.2</h2> <h2>🚀 urllib3 is fundraising for HTTP/2 support</h2> <p><a href="https://sethmlarson.dev/urllib3-is-fundraising-for-http2-support">urllib3 is raising ~$40,000 USD</a> to release HTTP/2 support and ensure long-term sustainable maintenance of the project after a sharp decline in financial support for 2023. If your company or organization uses Python and would benefit from HTTP/2 support in Requests, pip, cloud SDKs, and thousands of other projects <a href="https://opencollective.com/urllib3">please consider contributing financially</a> to ensure HTTP/2 support is developed sustainably and maintained for the long-haul.</p> <p>Thank you for your support.</p> <h2>Changes</h2> <ul> <li>Added the <code>Proxy-Authorization</code> header to the list of headers to strip from requests when redirecting to a different host. As before, different headers can be set via <code>Retry.remove_headers_on_redirect</code>.</li> <li>Allowed passing negative integers as <code>amt</code> to read methods of <code>http.client.HTTPResponse</code> as an alternative to <code>None</code>. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3122">#3122</a>)</li> <li>Fixed return types representing copying actions to use <code>typing.Self</code>. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3363">#3363</a>)</li> </ul> <p><strong>Full Changelog</strong>: <a href="https://github.com/urllib3/urllib3/compare/2.2.1...2.2.2">https://github.com/urllib3/urllib3/compare/2.2.1...2.2.2</a></p> <h2>2.2.1</h2> <h2>🚀 urllib3 is fundraising for HTTP/2 support</h2> <p><a href="https://sethmlarson.dev/urllib3-is-fundraising-for-http2-support">urllib3 is raising ~$40,000 USD</a> to release HTTP/2 support and ensure long-term sustainable maintenance of the project after a sharp decline in financial support for 2023. If your company or organization uses Python and would benefit from HTTP/2 support in Requests, pip, cloud SDKs, and thousands of other projects <a href="https://opencollective.com/urllib3">please consider contributing financially</a> to ensure HTTP/2 support is developed sustainably and maintained for the long-haul.</p> <p>Thank you for your support.</p> <h2>Changes</h2> <ul> <li>Fixed issue where <code>InsecureRequestWarning</code> was emitted for HTTPS connections when using Emscripten. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3331">#3331</a>)</li> <li>Fixed <code>HTTPConnectionPool.urlopen</code> to stop automatically casting non-proxy headers to <code>HTTPHeaderDict</code>. This change was premature as it did not apply to proxy headers and <code>HTTPHeaderDict</code> does not handle byte header values correctly yet. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3343">#3343</a>)</li> <li>Changed <code>ProtocolError</code> to <code>InvalidChunkLength</code> when response terminates before the chunk length is sent. (<a href="https://redirect.github.com/urllib3/urllib3/issues/2860">#2860</a>)</li> <li>Changed <code>ProtocolError</code> to be more verbose on incomplete reads with excess content. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3261">#3261</a>)</li> </ul> <h2>2.2.0</h2> <h2>🖥️ urllib3 now works in the browser</h2> <p>:tada: <strong>This release adds experimental support for <a href="https://urllib3.readthedocs.io/en/stable/reference/contrib/emscripten.html">using urllib3 in the browser with Pyodide</a>!</strong> :tada:</p> <p>Thanks to Joe Marshall (<a href="https://github.com/joemarshall"><code>@​joemarshall</code></a>) for contributing this feature. This change was possible thanks to work done in urllib3 v2.0 to detach our API from <code>http.client</code>. Please report all bugs to the <a href="https://github.com/urllib3/urllib3/issues">urllib3 issue tracker</a>.</p> <h2>🚀 urllib3 is fundraising for HTTP/2 support</h2> <p><a href="https://sethmlarson.dev/urllib3-is-fundraising-for-http2-support">urllib3 is raising ~$40,000 USD</a> to release HTTP/2 support and ensure long-term sustainable maintenance of the project after a sharp decline in financial support for 2023. If your company or organization uses Python and would benefit from HTTP/2 support in Requests, pip, cloud SDKs, and thousands of other projects <a href="https://opencollective.com/urllib3">please consider contributing financially</a> to ensure HTTP/2 support is developed sustainably and maintained for the long-haul.</p> <p>Thank you for your support.</p> <h2>Changes</h2> <ul> <li>Added support for <a href="https://urllib3.readthedocs.io/en/latest/reference/contrib/emscripten.html">Emscripten and Pyodide</a>, including streaming support in cross-origin isolated browser environments where threading is enabled. (<a href="https://redirect.github.com/urllib3/urllib3/issues/2951">#2951</a>)</li> <li>Added support for <code>HTTPResponse.read1()</code> method. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3186">#3186</a>)</li> <li>Added rudimentary support for HTTP/2. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3284">#3284</a>)</li> <li>Fixed issue where requests against urls with trailing dots were failing due to SSL errors when using proxy. (<a href="https://redirect.github.com/urllib3/urllib3/issues/2244">#2244</a>)</li> <li>Fixed <code>HTTPConnection.proxy_is_verified</code> and <code>HTTPSConnection.proxy_is_verified</code> to be always set to a boolean after connecting to a proxy. It could be <code>None</code> in some cases previously. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3130">#3130</a>)</li> </ul> <!-- raw HTML omitted --> </blockquote> <p>... (truncated)</p> </details> <details> <summary>Changelog</summary> <p><em>Sourced from <a href="https://github.com/urllib3/urllib3/blob/main/CHANGES.rst">urllib3's changelog</a>.</em></p> <blockquote> <h1>2.2.2 (2024-06-17)</h1> <ul> <li>Added the <code>Proxy-Authorization</code> header to the list of headers to strip from requests when redirecting to a different host. As before, different headers can be set via <code>Retry.remove_headers_on_redirect</code>.</li> <li>Allowed passing negative integers as <code>amt</code> to read methods of <code>http.client.HTTPResponse</code> as an alternative to <code>None</code>. (<code>[#3122](https://github.com/urllib3/urllib3/issues/3122) &lt;https://github.com/urllib3/urllib3/issues/3122&gt;</code>__)</li> <li>Fixed return types representing copying actions to use <code>typing.Self</code>. (<code>[#3363](https://github.com/urllib3/urllib3/issues/3363) &lt;https://github.com/urllib3/urllib3/issues/3363&gt;</code>__)</li> </ul> <h1>2.2.1 (2024-02-16)</h1> <ul> <li>Fixed issue where <code>InsecureRequestWarning</code> was emitted for HTTPS connections when using Emscripten. (<code>[#3331](https://github.com/urllib3/urllib3/issues/3331) &lt;https://github.com/urllib3/urllib3/issues/3331&gt;</code>__)</li> <li>Fixed <code>HTTPConnectionPool.urlopen</code> to stop automatically casting non-proxy headers to <code>HTTPHeaderDict</code>. This change was premature as it did not apply to proxy headers and <code>HTTPHeaderDict</code> does not handle byte header values correctly yet. (<code>[#3343](https://github.com/urllib3/urllib3/issues/3343) &lt;https://github.com/urllib3/urllib3/issues/3343&gt;</code>__)</li> <li>Changed <code>InvalidChunkLength</code> to <code>ProtocolError</code> when response terminates before the chunk length is sent. (<code>[#2860](https://github.com/urllib3/urllib3/issues/2860) &lt;https://github.com/urllib3/urllib3/issues/2860&gt;</code>__)</li> <li>Changed <code>ProtocolError</code> to be more verbose on incomplete reads with excess content. (<code>[#3261](https://github.com/urllib3/urllib3/issues/3261) &lt;https://github.com/urllib3/urllib3/issues/3261&gt;</code>__)</li> </ul> <h1>2.2.0 (2024-01-30)</h1> <ul> <li>Added support for <code>Emscripten and Pyodide &lt;https://urllib3.readthedocs.io/en/latest/reference/contrib/emscripten.html&gt;</code><strong>, including streaming support in cross-origin isolated browser environments where threading is enabled. (<code>[#2951](https://github.com/urllib3/urllib3/issues/2951) &lt;https://github.com/urllib3/urllib3/issues/2951&gt;</code></strong>)</li> <li>Added support for <code>HTTPResponse.read1()</code> method. (<code>[#3186](https://github.com/urllib3/urllib3/issues/3186) &lt;https://github.com/urllib3/urllib3/issues/3186&gt;</code>__)</li> <li>Added rudimentary support for HTTP/2. (<code>[#3284](https://github.com/urllib3/urllib3/issues/3284) &lt;https://github.com/urllib3/urllib3/issues/3284&gt;</code>__)</li> <li>Fixed issue where requests against urls with trailing dots were failing due to SSL errors when using proxy. (<code>[#2244](https://github.com/urllib3/urllib3/issues/2244) &lt;https://github.com/urllib3/urllib3/issues/2244&gt;</code>__)</li> <li>Fixed <code>HTTPConnection.proxy_is_verified</code> and <code>HTTPSConnection.proxy_is_verified</code> to be always set to a boolean after connecting to a proxy. It could be <code>None</code> in some cases previously. (<code>[#3130](https://github.com/urllib3/urllib3/issues/3130) &lt;https://github.com/urllib3/urllib3/issues/3130&gt;</code>__)</li> <li>Fixed an issue where <code>headers</code> passed in a request with <code>json=</code> would be mutated (<code>[#3203](https://github.com/urllib3/urllib3/issues/3203) &lt;https://github.com/urllib3/urllib3/issues/3203&gt;</code>__)</li> <li>Fixed <code>HTTPSConnection.is_verified</code> to be set to <code>False</code> when connecting from a HTTPS proxy to an HTTP target. It was set to <code>True</code> previously. (<code>[#3267](https://github.com/urllib3/urllib3/issues/3267) &lt;https://github.com/urllib3/urllib3/issues/3267&gt;</code>__)</li> <li>Fixed handling of new error message from OpenSSL 3.2.0 when configuring an HTTP proxy as HTTPS (<code>[#3268](https://github.com/urllib3/urllib3/issues/3268) &lt;https://github.com/urllib3/urllib3/issues/3268&gt;</code>__)</li> <li>Fixed TLS 1.3 post-handshake auth when the server certificate validation is disabled (<code>[#3325](https://github.com/urllib3/urllib3/issues/3325) &lt;https://github.com/urllib3/urllib3/issues/3325&gt;</code>__)</li> <li>Note for downstream distributors: To run integration tests, you now need to run the tests a second time with the <code>--integration</code> pytest flag. (<code>[#3181](https://github.com/urllib3/urllib3/issues/3181) &lt;https://github.com/urllib3/urllib3/issues/3181&gt;</code>__)</li> </ul> <h1>2.1.0 (2023-11-13)</h1> <ul> <li>Removed support for the deprecated urllib3[secure] extra. (<code>[#2680](https://github.com/urllib3/urllib3/issues/2680) &lt;https://github.com/urllib3/urllib3/issues/2680&gt;</code>__)</li> <li>Removed support for the deprecated SecureTransport TLS implementation. (<code>[#2681](https://github.com/urllib3/urllib3/issues/2681) &lt;https://github.com/urllib3/urllib3/issues/2681&gt;</code>__)</li> <li>Removed support for the end-of-life Python 3.7. (<code>[#3143](https://github.com/urllib3/urllib3/issues/3143) &lt;https://github.com/urllib3/urllib3/issues/3143&gt;</code>__)</li> <li>Allowed loading CA certificates from memory for proxies. (<code>[#3065](https://github.com/urllib3/urllib3/issues/3065) &lt;https://github.com/urllib3/urllib3/issues/3065&gt;</code>__)</li> <li>Fixed decoding Gzip-encoded responses which specified <code>x-gzip</code> content-encoding. (<code>[#3174](https://github.com/urllib3/urllib3/issues/3174) &lt;https://github.com/urllib3/urllib3/issues/3174&gt;</code>__)</li> </ul> </blockquote> </details> <details> <summary>Commits</summary> <ul> <li><a href="https://github.com/urllib3/urllib3/commit/27e2a5c5a7ab6a517252cc8dcef3ffa6ffb8f61a"><code>27e2a5c</code></a> Release 2.2.2 (<a href="https://redirect.github.com/urllib3/urllib3/issues/3406">#3406</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/accff72ecc2f6cf5a76d9570198a93ac7c90270e"><code>accff72</code></a> Merge pull request from GHSA-34jh-p97f-mpxf</li> <li><a href="https://github.com/urllib3/urllib3/commit/34be4a57e59eb7365bcc37d52e9f8271b5b8d0d3"><code>34be4a5</code></a> Pin CFFI to a new release candidate instead of a Git commit (<a href="https://redirect.github.com/urllib3/urllib3/issues/3398">#3398</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/da410581b6b3df73da976b5ce5eb20a4bd030437"><code>da41058</code></a> Bump browser-actions/setup-chrome from 1.6.0 to 1.7.1 (<a href="https://redirect.github.com/urllib3/urllib3/issues/3399">#3399</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/b07a669bd970d69847801148286b726f0570b625"><code>b07a669</code></a> Bump github/codeql-action from 2.13.4 to 3.25.6 (<a href="https://redirect.github.com/urllib3/urllib3/issues/3396">#3396</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/b8589ec9f8c4da91511e601b632ac06af7e7c10e"><code>b8589ec</code></a> Measure coverage with v4 of artifact actions (<a href="https://redirect.github.com/urllib3/urllib3/issues/3394">#3394</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/f3bdc5585111429e22c81b5fb26c3ec164d98b81"><code>f3bdc55</code></a> Allow triggering CI manually (<a href="https://redirect.github.com/urllib3/urllib3/issues/3391">#3391</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/52392654b30183129cf3ec06010306f517d9c146"><code>5239265</code></a> Fix HTTP version in debug log (<a href="https://redirect.github.com/urllib3/urllib3/issues/3316">#3316</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/b34619f94ece0c40e691a5aaf1304953d88089de"><code>b34619f</code></a> Bump actions/checkout to 4.1.4 (<a href="https://redirect.github.com/urllib3/urllib3/issues/3387">#3387</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/9961d14de7c920091d42d42ed76d5d479b80064d"><code>9961d14</code></a> Bump browser-actions/setup-chrome from 1.5.0 to 1.6.0 (<a href="https://redirect.github.com/urllib3/urllib3/issues/3386">#3386</a>)</li> <li>Additional commits viewable in <a href="https://github.com/urllib3/urllib3/compare/2.0.7...2.2.2">compare view</a></li> </ul> </details> <br /> [](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores) Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting `@dependabot rebase`. [//]: # (dependabot-automerge-start) [//]: # (dependabot-automerge-end) --- <details> <summary>Dependabot commands and options</summary> <br /> You can trigger Dependabot actions by commenting on this PR: - `@dependabot rebase` will rebase this PR - `@dependabot recreate` will recreate this PR, overwriting any edits that have been made to it - `@dependabot merge` will merge this PR after your CI passes on it - `@dependabot squash and merge` will squash and merge this PR after your CI passes on it - `@dependabot cancel merge` will cancel a previously requested merge and block automerging - `@dependabot reopen` will reopen this PR if it is closed - `@dependabot close` will close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually - `@dependabot show <dependency name> ignore conditions` will show all of the ignore conditions of the specified dependency - `@dependabot ignore <dependency name> major version` will close this group update PR and stop Dependabot creating any more for the specific dependency's major version (unless you unignore this specific dependency's major version or upgrade to it yourself) - `@dependabot ignore <dependency name> minor version` will close this group update PR and stop Dependabot creating any more for the specific dependency's minor version (unless you unignore this specific dependency's minor version or upgrade to it yourself) - `@dependabot ignore <dependency name>` will close this group update PR and stop Dependabot creating any more for the specific dependency (unless you unignore this specific dependency or upgrade to it yourself) - `@dependabot unignore <dependency name>` will remove all of the ignore conditions of the specified dependency - `@dependabot unignore <dependency name> <ignore condition>` will remove the ignore condition of the specified dependency and ignore conditions You can disable automated security fix PRs for this repo from the [Security Alerts page](https://github.com/huggingface/dataset-viewer/network/alerts). </details>

Bump urllib3 from 2.0.7 to 2.2.2 in /docs in the pip group across 1 directory: Bumps the pip group with 1 update in the /docs directory: [urllib3](https://github.com/urllib3/urllib3). Updates `urllib3` from 2.0.7 to 2.2.2 <details> <summary>Release notes</summary> <p><em>Sourced from <a href="https://github.com/urllib3/urllib3/releases">urllib3's releases</a>.</em></p> <blockquote> <h2>2.2.2</h2> <h2>🚀 urllib3 is fundraising for HTTP/2 support</h2> <p><a href="https://sethmlarson.dev/urllib3-is-fundraising-for-http2-support">urllib3 is raising ~$40,000 USD</a> to release HTTP/2 support and ensure long-term sustainable maintenance of the project after a sharp decline in financial support for 2023. If your company or organization uses Python and would benefit from HTTP/2 support in Requests, pip, cloud SDKs, and thousands of other projects <a href="https://opencollective.com/urllib3">please consider contributing financially</a> to ensure HTTP/2 support is developed sustainably and maintained for the long-haul.</p> <p>Thank you for your support.</p> <h2>Changes</h2> <ul> <li>Added the <code>Proxy-Authorization</code> header to the list of headers to strip from requests when redirecting to a different host. As before, different headers can be set via <code>Retry.remove_headers_on_redirect</code>.</li> <li>Allowed passing negative integers as <code>amt</code> to read methods of <code>http.client.HTTPResponse</code> as an alternative to <code>None</code>. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3122">#3122</a>)</li> <li>Fixed return types representing copying actions to use <code>typing.Self</code>. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3363">#3363</a>)</li> </ul> <p><strong>Full Changelog</strong>: <a href="https://github.com/urllib3/urllib3/compare/2.2.1...2.2.2">https://github.com/urllib3/urllib3/compare/2.2.1...2.2.2</a></p> <h2>2.2.1</h2> <h2>🚀 urllib3 is fundraising for HTTP/2 support</h2> <p><a href="https://sethmlarson.dev/urllib3-is-fundraising-for-http2-support">urllib3 is raising ~$40,000 USD</a> to release HTTP/2 support and ensure long-term sustainable maintenance of the project after a sharp decline in financial support for 2023. If your company or organization uses Python and would benefit from HTTP/2 support in Requests, pip, cloud SDKs, and thousands of other projects <a href="https://opencollective.com/urllib3">please consider contributing financially</a> to ensure HTTP/2 support is developed sustainably and maintained for the long-haul.</p> <p>Thank you for your support.</p> <h2>Changes</h2> <ul> <li>Fixed issue where <code>InsecureRequestWarning</code> was emitted for HTTPS connections when using Emscripten. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3331">#3331</a>)</li> <li>Fixed <code>HTTPConnectionPool.urlopen</code> to stop automatically casting non-proxy headers to <code>HTTPHeaderDict</code>. This change was premature as it did not apply to proxy headers and <code>HTTPHeaderDict</code> does not handle byte header values correctly yet. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3343">#3343</a>)</li> <li>Changed <code>ProtocolError</code> to <code>InvalidChunkLength</code> when response terminates before the chunk length is sent. (<a href="https://redirect.github.com/urllib3/urllib3/issues/2860">#2860</a>)</li> <li>Changed <code>ProtocolError</code> to be more verbose on incomplete reads with excess content. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3261">#3261</a>)</li> </ul> <h2>2.2.0</h2> <h2>🖥️ urllib3 now works in the browser</h2> <p>:tada: <strong>This release adds experimental support for <a href="https://urllib3.readthedocs.io/en/stable/reference/contrib/emscripten.html">using urllib3 in the browser with Pyodide</a>!</strong> :tada:</p> <p>Thanks to Joe Marshall (<a href="https://github.com/joemarshall"><code>@​joemarshall</code></a>) for contributing this feature. This change was possible thanks to work done in urllib3 v2.0 to detach our API from <code>http.client</code>. Please report all bugs to the <a href="https://github.com/urllib3/urllib3/issues">urllib3 issue tracker</a>.</p> <h2>🚀 urllib3 is fundraising for HTTP/2 support</h2> <p><a href="https://sethmlarson.dev/urllib3-is-fundraising-for-http2-support">urllib3 is raising ~$40,000 USD</a> to release HTTP/2 support and ensure long-term sustainable maintenance of the project after a sharp decline in financial support for 2023. If your company or organization uses Python and would benefit from HTTP/2 support in Requests, pip, cloud SDKs, and thousands of other projects <a href="https://opencollective.com/urllib3">please consider contributing financially</a> to ensure HTTP/2 support is developed sustainably and maintained for the long-haul.</p> <p>Thank you for your support.</p> <h2>Changes</h2> <ul> <li>Added support for <a href="https://urllib3.readthedocs.io/en/latest/reference/contrib/emscripten.html">Emscripten and Pyodide</a>, including streaming support in cross-origin isolated browser environments where threading is enabled. (<a href="https://redirect.github.com/urllib3/urllib3/issues/2951">#2951</a>)</li> <li>Added support for <code>HTTPResponse.read1()</code> method. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3186">#3186</a>)</li> <li>Added rudimentary support for HTTP/2. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3284">#3284</a>)</li> <li>Fixed issue where requests against urls with trailing dots were failing due to SSL errors when using proxy. (<a href="https://redirect.github.com/urllib3/urllib3/issues/2244">#2244</a>)</li> <li>Fixed <code>HTTPConnection.proxy_is_verified</code> and <code>HTTPSConnection.proxy_is_verified</code> to be always set to a boolean after connecting to a proxy. It could be <code>None</code> in some cases previously. (<a href="https://redirect.github.com/urllib3/urllib3/issues/3130">#3130</a>)</li> </ul> <!-- raw HTML omitted --> </blockquote> <p>... (truncated)</p> </details> <details> <summary>Changelog</summary> <p><em>Sourced from <a href="https://github.com/urllib3/urllib3/blob/main/CHANGES.rst">urllib3's changelog</a>.</em></p> <blockquote> <h1>2.2.2 (2024-06-17)</h1> <ul> <li>Added the <code>Proxy-Authorization</code> header to the list of headers to strip from requests when redirecting to a different host. As before, different headers can be set via <code>Retry.remove_headers_on_redirect</code>.</li> <li>Allowed passing negative integers as <code>amt</code> to read methods of <code>http.client.HTTPResponse</code> as an alternative to <code>None</code>. (<code>[#3122](https://github.com/urllib3/urllib3/issues/3122) &lt;https://github.com/urllib3/urllib3/issues/3122&gt;</code>__)</li> <li>Fixed return types representing copying actions to use <code>typing.Self</code>. (<code>[#3363](https://github.com/urllib3/urllib3/issues/3363) &lt;https://github.com/urllib3/urllib3/issues/3363&gt;</code>__)</li> </ul> <h1>2.2.1 (2024-02-16)</h1> <ul> <li>Fixed issue where <code>InsecureRequestWarning</code> was emitted for HTTPS connections when using Emscripten. (<code>[#3331](https://github.com/urllib3/urllib3/issues/3331) &lt;https://github.com/urllib3/urllib3/issues/3331&gt;</code>__)</li> <li>Fixed <code>HTTPConnectionPool.urlopen</code> to stop automatically casting non-proxy headers to <code>HTTPHeaderDict</code>. This change was premature as it did not apply to proxy headers and <code>HTTPHeaderDict</code> does not handle byte header values correctly yet. (<code>[#3343](https://github.com/urllib3/urllib3/issues/3343) &lt;https://github.com/urllib3/urllib3/issues/3343&gt;</code>__)</li> <li>Changed <code>InvalidChunkLength</code> to <code>ProtocolError</code> when response terminates before the chunk length is sent. (<code>[#2860](https://github.com/urllib3/urllib3/issues/2860) &lt;https://github.com/urllib3/urllib3/issues/2860&gt;</code>__)</li> <li>Changed <code>ProtocolError</code> to be more verbose on incomplete reads with excess content. (<code>[#3261](https://github.com/urllib3/urllib3/issues/3261) &lt;https://github.com/urllib3/urllib3/issues/3261&gt;</code>__)</li> </ul> <h1>2.2.0 (2024-01-30)</h1> <ul> <li>Added support for <code>Emscripten and Pyodide &lt;https://urllib3.readthedocs.io/en/latest/reference/contrib/emscripten.html&gt;</code><strong>, including streaming support in cross-origin isolated browser environments where threading is enabled. (<code>[#2951](https://github.com/urllib3/urllib3/issues/2951) &lt;https://github.com/urllib3/urllib3/issues/2951&gt;</code></strong>)</li> <li>Added support for <code>HTTPResponse.read1()</code> method. (<code>[#3186](https://github.com/urllib3/urllib3/issues/3186) &lt;https://github.com/urllib3/urllib3/issues/3186&gt;</code>__)</li> <li>Added rudimentary support for HTTP/2. (<code>[#3284](https://github.com/urllib3/urllib3/issues/3284) &lt;https://github.com/urllib3/urllib3/issues/3284&gt;</code>__)</li> <li>Fixed issue where requests against urls with trailing dots were failing due to SSL errors when using proxy. (<code>[#2244](https://github.com/urllib3/urllib3/issues/2244) &lt;https://github.com/urllib3/urllib3/issues/2244&gt;</code>__)</li> <li>Fixed <code>HTTPConnection.proxy_is_verified</code> and <code>HTTPSConnection.proxy_is_verified</code> to be always set to a boolean after connecting to a proxy. It could be <code>None</code> in some cases previously. (<code>[#3130](https://github.com/urllib3/urllib3/issues/3130) &lt;https://github.com/urllib3/urllib3/issues/3130&gt;</code>__)</li> <li>Fixed an issue where <code>headers</code> passed in a request with <code>json=</code> would be mutated (<code>[#3203](https://github.com/urllib3/urllib3/issues/3203) &lt;https://github.com/urllib3/urllib3/issues/3203&gt;</code>__)</li> <li>Fixed <code>HTTPSConnection.is_verified</code> to be set to <code>False</code> when connecting from a HTTPS proxy to an HTTP target. It was set to <code>True</code> previously. (<code>[#3267](https://github.com/urllib3/urllib3/issues/3267) &lt;https://github.com/urllib3/urllib3/issues/3267&gt;</code>__)</li> <li>Fixed handling of new error message from OpenSSL 3.2.0 when configuring an HTTP proxy as HTTPS (<code>[#3268](https://github.com/urllib3/urllib3/issues/3268) &lt;https://github.com/urllib3/urllib3/issues/3268&gt;</code>__)</li> <li>Fixed TLS 1.3 post-handshake auth when the server certificate validation is disabled (<code>[#3325](https://github.com/urllib3/urllib3/issues/3325) &lt;https://github.com/urllib3/urllib3/issues/3325&gt;</code>__)</li> <li>Note for downstream distributors: To run integration tests, you now need to run the tests a second time with the <code>--integration</code> pytest flag. (<code>[#3181](https://github.com/urllib3/urllib3/issues/3181) &lt;https://github.com/urllib3/urllib3/issues/3181&gt;</code>__)</li> </ul> <h1>2.1.0 (2023-11-13)</h1> <ul> <li>Removed support for the deprecated urllib3[secure] extra. (<code>[#2680](https://github.com/urllib3/urllib3/issues/2680) &lt;https://github.com/urllib3/urllib3/issues/2680&gt;</code>__)</li> <li>Removed support for the deprecated SecureTransport TLS implementation. (<code>[#2681](https://github.com/urllib3/urllib3/issues/2681) &lt;https://github.com/urllib3/urllib3/issues/2681&gt;</code>__)</li> <li>Removed support for the end-of-life Python 3.7. (<code>[#3143](https://github.com/urllib3/urllib3/issues/3143) &lt;https://github.com/urllib3/urllib3/issues/3143&gt;</code>__)</li> <li>Allowed loading CA certificates from memory for proxies. (<code>[#3065](https://github.com/urllib3/urllib3/issues/3065) &lt;https://github.com/urllib3/urllib3/issues/3065&gt;</code>__)</li> <li>Fixed decoding Gzip-encoded responses which specified <code>x-gzip</code> content-encoding. (<code>[#3174](https://github.com/urllib3/urllib3/issues/3174) &lt;https://github.com/urllib3/urllib3/issues/3174&gt;</code>__)</li> </ul> </blockquote> </details> <details> <summary>Commits</summary> <ul> <li><a href="https://github.com/urllib3/urllib3/commit/27e2a5c5a7ab6a517252cc8dcef3ffa6ffb8f61a"><code>27e2a5c</code></a> Release 2.2.2 (<a href="https://redirect.github.com/urllib3/urllib3/issues/3406">#3406</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/accff72ecc2f6cf5a76d9570198a93ac7c90270e"><code>accff72</code></a> Merge pull request from GHSA-34jh-p97f-mpxf</li> <li><a href="https://github.com/urllib3/urllib3/commit/34be4a57e59eb7365bcc37d52e9f8271b5b8d0d3"><code>34be4a5</code></a> Pin CFFI to a new release candidate instead of a Git commit (<a href="https://redirect.github.com/urllib3/urllib3/issues/3398">#3398</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/da410581b6b3df73da976b5ce5eb20a4bd030437"><code>da41058</code></a> Bump browser-actions/setup-chrome from 1.6.0 to 1.7.1 (<a href="https://redirect.github.com/urllib3/urllib3/issues/3399">#3399</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/b07a669bd970d69847801148286b726f0570b625"><code>b07a669</code></a> Bump github/codeql-action from 2.13.4 to 3.25.6 (<a href="https://redirect.github.com/urllib3/urllib3/issues/3396">#3396</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/b8589ec9f8c4da91511e601b632ac06af7e7c10e"><code>b8589ec</code></a> Measure coverage with v4 of artifact actions (<a href="https://redirect.github.com/urllib3/urllib3/issues/3394">#3394</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/f3bdc5585111429e22c81b5fb26c3ec164d98b81"><code>f3bdc55</code></a> Allow triggering CI manually (<a href="https://redirect.github.com/urllib3/urllib3/issues/3391">#3391</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/52392654b30183129cf3ec06010306f517d9c146"><code>5239265</code></a> Fix HTTP version in debug log (<a href="https://redirect.github.com/urllib3/urllib3/issues/3316">#3316</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/b34619f94ece0c40e691a5aaf1304953d88089de"><code>b34619f</code></a> Bump actions/checkout to 4.1.4 (<a href="https://redirect.github.com/urllib3/urllib3/issues/3387">#3387</a>)</li> <li><a href="https://github.com/urllib3/urllib3/commit/9961d14de7c920091d42d42ed76d5d479b80064d"><code>9961d14</code></a> Bump browser-actions/setup-chrome from 1.5.0 to 1.6.0 (<a href="https://redirect.github.com/urllib3/urllib3/issues/3386">#3386</a>)</li> <li>Additional commits viewable in <a href="https://github.com/urllib3/urllib3/compare/2.0.7...2.2.2">compare view</a></li> </ul> </details> <br /> [](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores) Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting `@dependabot rebase`. [//]: # (dependabot-automerge-start) [//]: # (dependabot-automerge-end) --- <details> <summary>Dependabot commands and options</summary> <br /> You can trigger Dependabot actions by commenting on this PR: - `@dependabot rebase` will rebase this PR - `@dependabot recreate` will recreate this PR, overwriting any edits that have been made to it - `@dependabot merge` will merge this PR after your CI passes on it - `@dependabot squash and merge` will squash and merge this PR after your CI passes on it - `@dependabot cancel merge` will cancel a previously requested merge and block automerging - `@dependabot reopen` will reopen this PR if it is closed - `@dependabot close` will close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually - `@dependabot show <dependency name> ignore conditions` will show all of the ignore conditions of the specified dependency - `@dependabot ignore <dependency name> major version` will close this group update PR and stop Dependabot creating any more for the specific dependency's major version (unless you unignore this specific dependency's major version or upgrade to it yourself) - `@dependabot ignore <dependency name> minor version` will close this group update PR and stop Dependabot creating any more for the specific dependency's minor version (unless you unignore this specific dependency's minor version or upgrade to it yourself) - `@dependabot ignore <dependency name>` will close this group update PR and stop Dependabot creating any more for the specific dependency (unless you unignore this specific dependency or upgrade to it yourself) - `@dependabot unignore <dependency name>` will remove all of the ignore conditions of the specified dependency - `@dependabot unignore <dependency name> <ignore condition>` will remove the ignore condition of the specified dependency and ignore conditions You can disable automated security fix PRs for this repo from the [Security Alerts page](https://github.com/huggingface/dataset-viewer/network/alerts). </details>

closed

2024-06-17T22:17:57Z

2024-07-27T15:04:09Z

2024-07-27T15:04:01Z

dependabot[bot]

Prevents viewer from being pinged for the datasets on both leaderboard orgs

Prevents viewer from being pinged for the datasets on both leaderboard orgs:

closed

2024-06-17T09:05:22Z

2024-06-17T09:06:18Z

2024-06-17T09:06:17Z

clefourrier

dataset-filetypes is not a small step

It has to do HTTP requests, it can manage big objects, and it depends on `datasets`. So... it might be better to process it with "medium" workers, not "light" ones that generally only transform cached entries.

dataset-filetypes is not a small step: It has to do HTTP requests, it can manage big objects, and it depends on `datasets`. So... it might be better to process it with "medium" workers, not "light" ones that generally only transform cached entries.

closed

2024-06-14T16:43:33Z

2024-06-14T17:05:45Z

2024-06-14T17:05:44Z

severo

Additional modalities detection

- [x] detect tabular - [x] detect timeseries

Additional modalities detection: - [x] detect tabular - [x] detect timeseries

closed

2024-06-14T13:58:17Z

2024-06-14T17:05:56Z

2024-06-14T17:05:56Z

severo

feat(chart): auto deploy when secrets change

Will automatically redeploy applications when secrets are changed in Infisical (max 1 min after the change).

feat(chart): auto deploy when secrets change: Will automatically redeploy applications when secrets are changed in Infisical (max 1 min after the change).

closed

2024-06-14T08:16:54Z

2024-06-26T08:20:17Z

2024-06-26T08:20:16Z

rtrompier

fix: extensions are always lowercase

followup to #2905

fix: extensions are always lowercase: followup to #2905

closed

2024-06-13T16:41:28Z

2024-06-13T20:48:49Z

2024-06-13T20:48:48Z

severo

Detect dataset modalities using dataset-filetypes

See #2898

Detect dataset modalities using dataset-filetypes: See #2898

closed

2024-06-13T15:34:21Z

2024-06-14T17:18:51Z

2024-06-14T17:18:50Z

severo

Add `started_at` field to cached response documents

Will close https://github.com/huggingface/dataset-viewer/issues/2892 To pass `started_at` field from Job to CachedResponse, I updated `JobInfo` class so that it stores `started_at` info too. It is None by default (and set to actual time by `Queue._start_newest_job_and_delete_others()`). Maybe there is better way to pass this information? I think alternatively I can get a new timestamp inside worker's `Loop.process_next_job` and pass it to `finish` to avoid changing dtos, but I don't know if it's better.

Add `started_at` field to cached response documents: Will close https://github.com/huggingface/dataset-viewer/issues/2892 To pass `started_at` field from Job to CachedResponse, I updated `JobInfo` class so that it stores `started_at` info too. It is None by default (and set to actual time by `Queue._start_newest_job_and_delete_others()`). Maybe there is better way to pass this information? I think alternatively I can get a new timestamp inside worker's `Loop.process_next_job` and pass it to `finish` to avoid changing dtos, but I don't know if it's better.

closed

2024-06-13T13:51:53Z

2024-07-04T15:29:08Z

2024-07-04T15:29:08Z

polinaeterna

Move secrets to Infisical

Move secrets to Infisical:

closed

2024-06-12T15:44:59Z

2024-06-13T18:17:14Z

2024-06-13T18:17:14Z

rtrompier

[Config-parquet-and-info] Compute estimated dataset info

This will be useful to show the estimate number of rows of datasets that are partially converted to Parquet I added `estimated_dataset_info` to the` parquet-and-info` response. It contains estimations of: - download_size - num_bytes - num_examples Then we'll be able to propagate this info to the `size` jobs and then to `hub-cache`. I'll run it on some datasets to check if it works fine and when it's ok I'll re-run the jobs for datasets for which we need to estimate the number of rows. TODO: - [x] compute sizes of non-read files - [x] test with compressed files - [x] test with zip archives - [x] add track_read() test (will add before merging)

[Config-parquet-and-info] Compute estimated dataset info: This will be useful to show the estimate number of rows of datasets that are partially converted to Parquet I added `estimated_dataset_info` to the` parquet-and-info` response. It contains estimations of: - download_size - num_bytes - num_examples Then we'll be able to propagate this info to the `size` jobs and then to `hub-cache`. I'll run it on some datasets to check if it works fine and when it's ok I'll re-run the jobs for datasets for which we need to estimate the number of rows. TODO: - [x] compute sizes of non-read files - [x] test with compressed files - [x] test with zip archives - [x] add track_read() test (will add before merging)

closed

2024-06-12T15:31:57Z

2024-06-19T13:57:05Z

2024-06-19T13:57:04Z

lhoestq

Add step dataset-filetypes

needed for https://github.com/huggingface/dataset-viewer/issues/2898#issuecomment-2162460762

Add step dataset-filetypes: needed for https://github.com/huggingface/dataset-viewer/issues/2898#issuecomment-2162460762

closed

2024-06-12T09:37:27Z

2024-06-13T15:13:07Z

2024-06-13T15:13:06Z

severo

Fix skipped async tests caused by pytest-memray

Fix skipped async tests caused by pytest-memray: do not pass memray argument to pytest. Fix #2901. Reported underlying issue in pytest-memray: - https://github.com/bloomberg/pytest-memray/issues/119

Fix skipped async tests caused by pytest-memray: Fix skipped async tests caused by pytest-memray: do not pass memray argument to pytest. Fix #2901. Reported underlying issue in pytest-memray: - https://github.com/bloomberg/pytest-memray/issues/119

closed

2024-06-12T08:46:37Z

2024-06-17T13:19:44Z

2024-06-12T11:25:18Z

albertvillanova

Pass copies of DataFrames instead of views

As the memory leak may be caused by improperly de-referenced objects, better pass copies of DataFrames instead of views. In a subsequent PR I could try to optimize memory usage by not storing unnecessary data.

Pass copies of DataFrames instead of views: As the memory leak may be caused by improperly de-referenced objects, better pass copies of DataFrames instead of views. In a subsequent PR I could try to optimize memory usage by not storing unnecessary data.

closed

2024-06-12T08:37:00Z

2024-06-20T07:59:38Z

2024-06-20T07:59:38Z

albertvillanova

Minor fix id with length of str dataset name

This is a minor fix of some Tasks ids containing the length of of the string dataset name. I discovered this while investigating the memory leak issue.

Minor fix id with length of str dataset name: This is a minor fix of some Tasks ids containing the length of of the string dataset name. I discovered this while investigating the memory leak issue.

closed

2024-06-12T06:30:18Z

2024-06-12T08:04:27Z

2024-06-12T08:03:44Z

albertvillanova

Async tests using anyio are skipped after including pytest-memray

Async tests using anyio are currently skipped. See: https://github.com/huggingface/dataset-viewer/actions/runs/9464138411/job/26070809625 ``` tests/test_authentication.py ssssssssssssssssssssssssssssssssssssssssssssss =============================== warnings summary =============================== /home/runner/.cache/pypoetry/virtualenvs/libapi-QfqNi3gs-py3.9/lib/python3.9/site-packages/_pytest/python.py:151: PytestUnhandledCoroutineWarning: async def functions are not natively supported and have been skipped. You need to install a suitable plugin for your async framework, for example: - anyio - pytest-asyncio - pytest-tornasync - pytest-trio - pytest-twisted warnings.warn(PytestUnhandledCoroutineWarning(msg.format(nodeid))) tests/test_authentication.py::test_no_external_auth_check /home/runner/.cache/pypoetry/virtualenvs/libapi-QfqNi3gs-py3.9/lib/python3.9/site-packages/pymongo/topology.py:498: RuntimeWarning: coroutine 'test_no_external_auth_check' was never awaited ... ``` After investigation, I found that the issue arose after including pytest-memray by PR: - https://github.com/huggingface/dataset-viewer/pull/2863 - The issue is caused when passing `--memray` to `pytest`: https://github.com/huggingface/dataset-viewer/blob/04f89d8a76e73874c9198c940ad2e61fe71f6f76/tools/PythonTest.mk#L7 I have reported the issue to pytest-memray team: - https://github.com/bloomberg/pytest-memray/discussions/101#discussioncomment-9738673 - https://github.com/bloomberg/pytest-memray/issues/119

Async tests using anyio are skipped after including pytest-memray: Async tests using anyio are currently skipped. See: https://github.com/huggingface/dataset-viewer/actions/runs/9464138411/job/26070809625 ``` tests/test_authentication.py ssssssssssssssssssssssssssssssssssssssssssssss =============================== warnings summary =============================== /home/runner/.cache/pypoetry/virtualenvs/libapi-QfqNi3gs-py3.9/lib/python3.9/site-packages/_pytest/python.py:151: PytestUnhandledCoroutineWarning: async def functions are not natively supported and have been skipped. You need to install a suitable plugin for your async framework, for example: - anyio - pytest-asyncio - pytest-tornasync - pytest-trio - pytest-twisted warnings.warn(PytestUnhandledCoroutineWarning(msg.format(nodeid))) tests/test_authentication.py::test_no_external_auth_check /home/runner/.cache/pypoetry/virtualenvs/libapi-QfqNi3gs-py3.9/lib/python3.9/site-packages/pymongo/topology.py:498: RuntimeWarning: coroutine 'test_no_external_auth_check' was never awaited ... ``` After investigation, I found that the issue arose after including pytest-memray by PR: - https://github.com/huggingface/dataset-viewer/pull/2863 - The issue is caused when passing `--memray` to `pytest`: https://github.com/huggingface/dataset-viewer/blob/04f89d8a76e73874c9198c940ad2e61fe71f6f76/tools/PythonTest.mk#L7 I have reported the issue to pytest-memray team: - https://github.com/bloomberg/pytest-memray/discussions/101#discussioncomment-9738673 - https://github.com/bloomberg/pytest-memray/issues/119

closed

2024-06-12T04:41:05Z

2024-06-12T11:25:19Z

2024-06-12T11:25:18Z

albertvillanova

2754 partial instead of error

fix #2754

2754 partial instead of error: fix #2754

closed

2024-06-11T14:40:02Z

2024-06-14T13:00:31Z

2024-06-13T13:57:19Z

severo

Standardize access to metrics and healthcheck

In some apps, the metrics and healthcheck are public: - https://datasets-server.huggingface.co/admin/metrics - https://datasets-server.huggingface.co/sse/metrics - https://datasets-server.huggingface.co/sse/healthcheck - https://datasets-server.huggingface.co/healthcheck - On others, it’s forbidden or not found: - https://datasets-server.huggingface.co/metrics - https://datasets-server.huggingface.co/filter/metrics As @severo suggests, it should be coherent among all the services. (Do we want the metrics to be public, or not?)

Standardize access to metrics and healthcheck: In some apps, the metrics and healthcheck are public: - https://datasets-server.huggingface.co/admin/metrics - https://datasets-server.huggingface.co/sse/metrics - https://datasets-server.huggingface.co/sse/healthcheck - https://datasets-server.huggingface.co/healthcheck - On others, it’s forbidden or not found: - https://datasets-server.huggingface.co/metrics - https://datasets-server.huggingface.co/filter/metrics As @severo suggests, it should be coherent among all the services. (Do we want the metrics to be public, or not?)

open

2024-06-11T14:39:10Z

2024-07-11T15:38:17Z

AndreaFrancis

detect more modalities

Currently, we only detect and report "audio", "image" and "text". Ideally, we would have: <img width="318" alt="Capture d’écran 2024-06-11 à 16 07 20" src="https://github.com/huggingface/dataset-viewer/assets/1676121/1a21dfff-5c78-45bd-8baf-de1f4b203b6a"> See https://github.com/huggingface/moon-landing/pull/9352#discussion_r1634909052 (internal) --- TODO: - [x] text - [x] image - [x] audio - [ ] video: detect extensions in the repo files: mp4, avi, mkv, etc. - [ ] 3d: detect extensions in the repo files: ply, glb, gltf, etc. - [ ] tabular <- I'm not sure what it means. Only numbers? - [ ] geospatial. Let's include both raster (e.g., satellite images) and vector (e.g. latitude/longitud points) even if they have few things in common and are generally in different datasets. Several strategies: extensions (geojson, geoparquet, geotiff, etc), column types (see geoarrow), column names? (longitude, latitude, geom), column contents (see wkt/wkb - but I don't think we want to look inside the data itself) - [ ] time-series: columns with type `Sequence(Value(float32))` which are not embeddings (filter against the column name: `emb`?) --- Currently, we only check the features given by the datasets library (`dataset-info` step). For the strategy based on looking at the file extensions, it's a bit imprecise (eg: if the YAML contains `data_files: data.csv` and the repo also contains a video, the dataset will be marked as video, while the video is ignored). We can bear with it for now, I think. Should we also detect repos that contain images or audio files but are not supported by the datasets library / the viewer?

detect more modalities: Currently, we only detect and report "audio", "image" and "text". Ideally, we would have: <img width="318" alt="Capture d’écran 2024-06-11 à 16 07 20" src="https://github.com/huggingface/dataset-viewer/assets/1676121/1a21dfff-5c78-45bd-8baf-de1f4b203b6a"> See https://github.com/huggingface/moon-landing/pull/9352#discussion_r1634909052 (internal) --- TODO: - [x] text - [x] image - [x] audio - [ ] video: detect extensions in the repo files: mp4, avi, mkv, etc. - [ ] 3d: detect extensions in the repo files: ply, glb, gltf, etc. - [ ] tabular <- I'm not sure what it means. Only numbers? - [ ] geospatial. Let's include both raster (e.g., satellite images) and vector (e.g. latitude/longitud points) even if they have few things in common and are generally in different datasets. Several strategies: extensions (geojson, geoparquet, geotiff, etc), column types (see geoarrow), column names? (longitude, latitude, geom), column contents (see wkt/wkb - but I don't think we want to look inside the data itself) - [ ] time-series: columns with type `Sequence(Value(float32))` which are not embeddings (filter against the column name: `emb`?) --- Currently, we only check the features given by the datasets library (`dataset-info` step). For the strategy based on looking at the file extensions, it's a bit imprecise (eg: if the YAML contains `data_files: data.csv` and the repo also contains a video, the dataset will be marked as video, while the video is ignored). We can bear with it for now, I think. Should we also detect repos that contain images or audio files but are not supported by the datasets library / the viewer?

closed

2024-06-11T14:07:26Z

2024-06-14T17:18:52Z

2024-06-14T17:18:51Z

severo

Use `HfFileSystem` in config-parquet-metadata step instead of `HttpFileSystem`

`config-parquet-metadata` step is failing again for [FineWeb](https://huggingface.co/datasets/HuggingFaceFW/fineweb) with errors like ``` "Could not read the parquet files: 504, message='Gateway Time-out', url=URL('https://huggingface.co/datasets/HuggingFaceFW/fineweb/resolve/refs%2Fconvert%2Fparquet/default/train-part1/4089.parquet')" ``` Maybe this would help (the same is used in `config-parquet-and-info` step which works). Previous fix was https://github.com/huggingface/dataset-viewer/pull/2884

Use `HfFileSystem` in config-parquet-metadata step instead of `HttpFileSystem`: `config-parquet-metadata` step is failing again for [FineWeb](https://huggingface.co/datasets/HuggingFaceFW/fineweb) with errors like ``` "Could not read the parquet files: 504, message='Gateway Time-out', url=URL('https://huggingface.co/datasets/HuggingFaceFW/fineweb/resolve/refs%2Fconvert%2Fparquet/default/train-part1/4089.parquet')" ``` Maybe this would help (the same is used in `config-parquet-and-info` step which works). Previous fix was https://github.com/huggingface/dataset-viewer/pull/2884

closed

2024-06-11T12:50:24Z

2024-06-13T16:09:14Z

2024-06-13T16:09:13Z

polinaeterna

Remove Prometheus context label

Remove Prometheus context label. Fix #2895.

Remove Prometheus context label: Remove Prometheus context label. Fix #2895.

closed

2024-06-11T09:18:13Z

2024-06-11T11:23:11Z

2024-06-11T10:44:09Z

albertvillanova

Too high label cardinality metrics in Prometheus

As discussed privately (CC: @McPatate), we may have too high label cardinality metrics in Prometheus. As stated in Prometheus docs: https://prometheus.io/docs/practices/naming/#labels > CAUTION: Remember that every unique combination of key-value label pairs represents a new time series, which can dramatically increase the amount of data stored. Do not use labels to store dimensions with high cardinality (many different label values), such as user IDs, email addresses, or other unbounded sets of values. Currently, our `StepProfiler` uses 3 labels: `method`, `step` and `context`. Recently, we fixed an issue with object memory addresses as values in the `context` label: - #2893 Note that sometimes we pass the dataset name in the `context` label, for example in `DatasetBackfillPlan`: https://github.com/huggingface/dataset-viewer/blob/19c049499af1b37cf4d9dfa93b58657d66b75337/libs/libcommon/src/libcommon/orchestrator.py#L490 Other times we pass the dataset and the config names in the `context` label, for example in `ConfigState`: https://github.com/huggingface/dataset-viewer/blob/19c049499af1b37cf4d9dfa93b58657d66b75337/libs/libcommon/src/libcommon/state.py#L202 For the (`method`, `step`, `context`) labels, this creates (number_of_method_values * number_of_step_values * **number_of_datasets**) or (number_of_method_values * number_of_step_values * **number_of_dataset_configs**) unique label values! This might be a too large cardinality. One possible solution to reduce cardinality would be to suppress the `context` label. What do you think? Other proposals? CC: @huggingface/dataset-viewer

Too high label cardinality metrics in Prometheus: As discussed privately (CC: @McPatate), we may have too high label cardinality metrics in Prometheus. As stated in Prometheus docs: https://prometheus.io/docs/practices/naming/#labels > CAUTION: Remember that every unique combination of key-value label pairs represents a new time series, which can dramatically increase the amount of data stored. Do not use labels to store dimensions with high cardinality (many different label values), such as user IDs, email addresses, or other unbounded sets of values. Currently, our `StepProfiler` uses 3 labels: `method`, `step` and `context`. Recently, we fixed an issue with object memory addresses as values in the `context` label: - #2893 Note that sometimes we pass the dataset name in the `context` label, for example in `DatasetBackfillPlan`: https://github.com/huggingface/dataset-viewer/blob/19c049499af1b37cf4d9dfa93b58657d66b75337/libs/libcommon/src/libcommon/orchestrator.py#L490 Other times we pass the dataset and the config names in the `context` label, for example in `ConfigState`: https://github.com/huggingface/dataset-viewer/blob/19c049499af1b37cf4d9dfa93b58657d66b75337/libs/libcommon/src/libcommon/state.py#L202 For the (`method`, `step`, `context`) labels, this creates (number_of_method_values * number_of_step_values * **number_of_datasets**) or (number_of_method_values * number_of_step_values * **number_of_dataset_configs**) unique label values! This might be a too large cardinality. One possible solution to reduce cardinality would be to suppress the `context` label. What do you think? Other proposals? CC: @huggingface/dataset-viewer

closed

2024-06-11T05:29:14Z

2024-06-11T10:44:10Z

2024-06-11T10:44:10Z

albertvillanova

feat(ci): add trufflehog secrets detection

### What does this PR do? Adding a GH action to scan for leaked secrets on each commit. ### Context `trufflehog` will scan the commit that triggered the CI for any token leak. `trufflehog` works with a large number of what they call "detectors", each of which will read the text from the commit to see if there is match for a token. For example, the hugging face detector will check for hf tokens and then query our `/api/whoami{-v2}` endpoint to check if the token is valid. If it detects a valid token, the CI will fail, informing you that you need to rotate the token given it leaked. ### References - https://github.com/trufflesecurity/trufflehog - https://github.com/marketplace/actions/trufflehog-oss

feat(ci): add trufflehog secrets detection: ### What does this PR do? Adding a GH action to scan for leaked secrets on each commit. ### Context `trufflehog` will scan the commit that triggered the CI for any token leak. `trufflehog` works with a large number of what they call "detectors", each of which will read the text from the commit to see if there is match for a token. For example, the hugging face detector will check for hf tokens and then query our `/api/whoami{-v2}` endpoint to check if the token is valid. If it detects a valid token, the CI will fail, informing you that you need to rotate the token given it leaked. ### References - https://github.com/trufflesecurity/trufflehog - https://github.com/marketplace/actions/trufflehog-oss

closed

2024-06-10T08:57:46Z

2024-06-10T09:18:37Z

2024-06-10T09:18:36Z

McPatate

Fix string representation of storage client

Fix string representation of storage client. The string representation of the storage client is used by the orchestrator `DeleteDatasetStorageTask` in both the task `id` attribute and as a `label` for prometheus_client Histogram: - https://github.com/huggingface/dataset-viewer/blob/1ea458cb22396f8af955bf5b1ebbb92d89f0a707/libs/libcommon/src/libcommon/orchestrator.py#L221 - https://github.com/huggingface/dataset-viewer/blob/1ea458cb22396f8af955bf5b1ebbb92d89f0a707/libs/libcommon/src/libcommon/orchestrator.py#L234 Since the inclusion of the URL signer to the storage client by PR: - #2298 the string representation of the storage client is something like: ``` StorageClient(protocol=s3, storage_root=hf-datasets-server-statics-test/assets, base_url=https://datasets-server-test.us.dev.moon.huggingface.tech/assets, overwrite=True), url_signer=<libcommon.cloudfront.CloudFront object at 0x55f27a209110>) ``` - See the url_signer: `url_signer=<libcommon.cloudfront.CloudFront object at 0x55f27a209110>` - This is a different value for each `CloudFront` instantiation This PR fixes the string representation of the URLSigner to be its class name, and therefore it fixes as well the string representation of the storage client to be something like: ``` StorageClient(protocol=s3, storage_root=hf-datasets-server-statics-test/assets, base_url=https://datasets-server-test.us.dev.moon.huggingface.tech/assets, overwrite=True), url_signer=CloudFront) ``` - See the url_signer is now: `url_signer=CloudFront`

Fix string representation of storage client: Fix string representation of storage client. The string representation of the storage client is used by the orchestrator `DeleteDatasetStorageTask` in both the task `id` attribute and as a `label` for prometheus_client Histogram: - https://github.com/huggingface/dataset-viewer/blob/1ea458cb22396f8af955bf5b1ebbb92d89f0a707/libs/libcommon/src/libcommon/orchestrator.py#L221 - https://github.com/huggingface/dataset-viewer/blob/1ea458cb22396f8af955bf5b1ebbb92d89f0a707/libs/libcommon/src/libcommon/orchestrator.py#L234 Since the inclusion of the URL signer to the storage client by PR: - #2298 the string representation of the storage client is something like: ``` StorageClient(protocol=s3, storage_root=hf-datasets-server-statics-test/assets, base_url=https://datasets-server-test.us.dev.moon.huggingface.tech/assets, overwrite=True), url_signer=<libcommon.cloudfront.CloudFront object at 0x55f27a209110>) ``` - See the url_signer: `url_signer=<libcommon.cloudfront.CloudFront object at 0x55f27a209110>` - This is a different value for each `CloudFront` instantiation This PR fixes the string representation of the URLSigner to be its class name, and therefore it fixes as well the string representation of the storage client to be something like: ``` StorageClient(protocol=s3, storage_root=hf-datasets-server-statics-test/assets, base_url=https://datasets-server-test.us.dev.moon.huggingface.tech/assets, overwrite=True), url_signer=CloudFront) ``` - See the url_signer is now: `url_signer=CloudFront`

closed

2024-06-10T08:40:59Z

2024-06-10T09:41:23Z

2024-06-10T09:41:22Z

albertvillanova

Store `started_at` or duration info in cached steps too

This would help to better monitor how much time a step take, especially depending on dataset size, so that we can see how certain changes influence processing speed and make more informed decisions about these changes and our size limits (see https://github.com/huggingface/dataset-viewer/issues/2878). Especially useful for duckdb and statistics step. I can work on this if you think it's useful (I think it is) and don't break any logic.

Store `started_at` or duration info in cached steps too: This would help to better monitor how much time a step take, especially depending on dataset size, so that we can see how certain changes influence processing speed and make more informed decisions about these changes and our size limits (see https://github.com/huggingface/dataset-viewer/issues/2878). Especially useful for duckdb and statistics step. I can work on this if you think it's useful (I think it is) and don't break any logic.

closed

2024-06-07T13:22:50Z

2024-07-09T13:06:36Z

2024-07-09T13:06:36Z

polinaeterna

Make StorageClient not warn when deleting a non-existing directory

Make StorageClient not warn when trying to delete a non-existing directory. I think we should only log a warning if the directory exists and could not be deleted.

Make StorageClient not warn when deleting a non-existing directory: Make StorageClient not warn when trying to delete a non-existing directory. I think we should only log a warning if the directory exists and could not be deleted.

closed

2024-06-07T07:12:26Z

2024-06-10T13:54:00Z

2024-06-10T13:53:59Z

albertvillanova

No mongo cache in DatasetRemovalPlan

don't keep the full query result in memory (as mongoengine does by default) this should reduce the frequency of memory spikes and could have an effect of the memory issues we're having

No mongo cache in DatasetRemovalPlan: don't keep the full query result in memory (as mongoengine does by default) this should reduce the frequency of memory spikes and could have an effect of the memory issues we're having

closed

2024-06-06T17:33:46Z

2024-06-07T11:17:46Z

2024-06-07T11:17:45Z

lhoestq

No auto backfill on most nfaa datasets

they correlate with memory spikes, I'm just hardcoding this to see the impact on memory and further investigate

No auto backfill on most nfaa datasets: they correlate with memory spikes, I'm just hardcoding this to see the impact on memory and further investigate

closed

2024-06-06T15:20:27Z

2024-06-06T15:59:12Z

2024-06-06T15:59:11Z

lhoestq

Fix get_shape in statistics when argument is bytes, not dict

Will fix duckdb-index step for [common-canvas/commoncatalog-cc-by](https://huggingface.co/datasets/common-canvas/commoncatalog-cc-by)

Fix get_shape in statistics when argument is bytes, not dict : Will fix duckdb-index step for [common-canvas/commoncatalog-cc-by](https://huggingface.co/datasets/common-canvas/commoncatalog-cc-by)

closed

2024-06-06T14:47:47Z

2024-06-06T16:58:07Z

2024-06-06T16:58:06Z

polinaeterna

Update ruff to 0.4.8

Update ruff to 0.4.8: https://github.com/astral-sh/ruff/releases/tag/v0.4.8 > Linter performance has been improved by around 10% on some microbenchmarks

Update ruff to 0.4.8: Update ruff to 0.4.8: https://github.com/astral-sh/ruff/releases/tag/v0.4.8 > Linter performance has been improved by around 10% on some microbenchmarks

closed

2024-06-06T04:41:04Z

2024-06-06T13:27:24Z

2024-06-06T10:16:06Z

albertvillanova

Update uvicorn (restart expired workers)

Update uvicorn (restart expired workers):

closed

2024-06-05T15:38:54Z

2024-06-06T10:27:10Z

2024-06-06T10:02:46Z

lhoestq

add missing deps to dev images

otherwise I can't build the images on mac m2 I didn't touch the prod images

add missing deps to dev images: otherwise I can't build the images on mac m2 I didn't touch the prod images

closed

2024-06-05T14:36:42Z

2024-06-06T13:28:48Z

2024-06-06T13:28:47Z

lhoestq

Add retry mechanism to get_parquet_file in parquet metadata step

...to see if it helps with [FineWeb](https://huggingface.co/datasets/HuggingFaceFW/fineweb) `config-parquet-metadata` issue. Currently the error says just `Server disconnected` which seems to be a `aiohttp.ServerDisconnectedError` error. If that works, a more fundamental solution would be to completely switch to `HfFyleSystem` instead of `HTTPFileSystem` and remove retries.

Add retry mechanism to get_parquet_file in parquet metadata step: ...to see if it helps with [FineWeb](https://huggingface.co/datasets/HuggingFaceFW/fineweb) `config-parquet-metadata` issue. Currently the error says just `Server disconnected` which seems to be a `aiohttp.ServerDisconnectedError` error. If that works, a more fundamental solution would be to completely switch to `HfFyleSystem` instead of `HTTPFileSystem` and remove retries.

closed

2024-06-05T12:41:23Z

2024-06-05T15:06:59Z

2024-06-05T13:25:30Z

polinaeterna

Update pytest to 8.2.2 and pytest-asyncio to 0.23.7

Update pytest to 8.2.2: https://github.com/pytest-dev/pytest/releases/tag/8.2.2 The update of pytest requires the update of pytest-asyncio as well. See: - https://github.com/pytest-dev/pytest-asyncio/pull/823 Otherwise, we get an AttributeError: https://github.com/huggingface/dataset-viewer/actions/runs/9378353611/job/25821435791 ``` AttributeError: 'FixtureDef' object has no attribute 'unittest' ``` ``` ==================================== ERRORS ==================================== ____________________ ERROR at setup of test_get_healthcheck ____________________ event_loop = <_UnixSelectorEventLoop running=False closed=False debug=False> request = <SubRequest 'app_test' for <Function test_get_healthcheck>> kwargs = {'app_config': AppConfig(api=ApiConfig(external_auth_url='http://localhost:8888/api/datasets/%s/auth-check', hf_auth_p...onfig(level=20), queue=QueueConfig(mongo_database='dataset_viewer_queue_test', mongo_url='mongodb://localhost:27017'))} @functools.wraps(fixture) def _asyncgen_fixture_wrapper( event_loop: asyncio.AbstractEventLoop, request: SubRequest, **kwargs: Any ): func = _perhaps_rebind_fixture_func( > fixture, request.instance, fixturedef.unittest ) E AttributeError: 'FixtureDef' object has no attribute 'unittest' /home/runner/.cache/pypoetry/virtualenvs/sse-api-Z9Qw0xoN-py3.9/lib/python3.9/site-packages/pytest_asyncio/plugin.py:276: AttributeError =========================== short test summary info ============================ ERROR tests/test_app.py::test_get_healthcheck - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_metrics - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_only_updates - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_only_initialization[?all=true-expected_events0] - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_only_initialization[-expected_events1] - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_only_initialization[?all=false-expected_events2] - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_initialization_and_updates[?all=true-expected_events0] - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_initialization_and_updates[?all=false-expected_events1] - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_initialization_and_updates[-expected_events2] - AttributeError: 'FixtureDef' object has no attribute 'unittest' ========================= 1 passed, 9 errors in 1.09s ========================== ```

Update pytest to 8.2.2 and pytest-asyncio to 0.23.7: Update pytest to 8.2.2: https://github.com/pytest-dev/pytest/releases/tag/8.2.2 The update of pytest requires the update of pytest-asyncio as well. See: - https://github.com/pytest-dev/pytest-asyncio/pull/823 Otherwise, we get an AttributeError: https://github.com/huggingface/dataset-viewer/actions/runs/9378353611/job/25821435791 ``` AttributeError: 'FixtureDef' object has no attribute 'unittest' ``` ``` ==================================== ERRORS ==================================== ____________________ ERROR at setup of test_get_healthcheck ____________________ event_loop = <_UnixSelectorEventLoop running=False closed=False debug=False> request = <SubRequest 'app_test' for <Function test_get_healthcheck>> kwargs = {'app_config': AppConfig(api=ApiConfig(external_auth_url='http://localhost:8888/api/datasets/%s/auth-check', hf_auth_p...onfig(level=20), queue=QueueConfig(mongo_database='dataset_viewer_queue_test', mongo_url='mongodb://localhost:27017'))} @functools.wraps(fixture) def _asyncgen_fixture_wrapper( event_loop: asyncio.AbstractEventLoop, request: SubRequest, **kwargs: Any ): func = _perhaps_rebind_fixture_func( > fixture, request.instance, fixturedef.unittest ) E AttributeError: 'FixtureDef' object has no attribute 'unittest' /home/runner/.cache/pypoetry/virtualenvs/sse-api-Z9Qw0xoN-py3.9/lib/python3.9/site-packages/pytest_asyncio/plugin.py:276: AttributeError =========================== short test summary info ============================ ERROR tests/test_app.py::test_get_healthcheck - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_metrics - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_only_updates - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_only_initialization[?all=true-expected_events0] - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_only_initialization[-expected_events1] - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_only_initialization[?all=false-expected_events2] - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_initialization_and_updates[?all=true-expected_events0] - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_initialization_and_updates[?all=false-expected_events1] - AttributeError: 'FixtureDef' object has no attribute 'unittest' ERROR tests/test_app.py::test_hub_cache_initialization_and_updates[-expected_events2] - AttributeError: 'FixtureDef' object has no attribute 'unittest' ========================= 1 passed, 9 errors in 1.09s ========================== ```

closed

2024-06-05T04:48:39Z

2024-06-06T10:42:21Z

2024-06-06T10:42:20Z

albertvillanova

Apply recommendations from duckdb to improve speed

DuckDB has a dedicated page called "My Workload Is Slow" https://duckdb.org/docs/guides/performance/my_workload_is_slow and more generally all the https://duckdb.org/docs/guides/performance/overview section. It could be good to review if some recommendations apply to our usage of duckdb.

Apply recommendations from duckdb to improve speed: DuckDB has a dedicated page called "My Workload Is Slow" https://duckdb.org/docs/guides/performance/my_workload_is_slow and more generally all the https://duckdb.org/docs/guides/performance/overview section. It could be good to review if some recommendations apply to our usage of duckdb.

open

2024-06-04T13:24:33Z

2024-06-04T13:48:48Z

severo

Remove canonical datasets from docs

Remove canonical datasets from docs, now that we no longer have canonical datasets.

Remove canonical datasets from docs: Remove canonical datasets from docs, now that we no longer have canonical datasets.

closed

2024-06-04T12:22:43Z

2024-07-08T06:34:01Z

2024-07-08T06:34:01Z

albertvillanova

Allow mnist and fashion mnist + remove canonical dataset logic

Allow mnist and fashion mnist + remove canonical dataset logic:

closed

2024-06-04T12:13:35Z

2024-06-04T13:16:32Z

2024-06-04T13:16:32Z

lhoestq

Use pymongoarrow to get dataset results as dataframe

Fix for https://github.com/huggingface/dataset-viewer/issues/2868

Use pymongoarrow to get dataset results as dataframe: Fix for https://github.com/huggingface/dataset-viewer/issues/2868

closed

2024-06-03T20:29:48Z

2024-06-05T13:32:08Z

2024-06-05T13:32:07Z

AndreaFrancis

Remove or increase the 5GB limit?

The dataset viewer shows statistics and provides filter + sort + search only for the first 5GB of each split. We are also unable to provide the exact number of rows for bigger splits. Note that we "show" all the rows for parquet-native datasets (i.e., we can access the rows randomly, i.e., we have pagination). Should we provide a way to increase or remove this limit?

Remove or increase the 5GB limit?: The dataset viewer shows statistics and provides filter + sort + search only for the first 5GB of each split. We are also unable to provide the exact number of rows for bigger splits. Note that we "show" all the rows for parquet-native datasets (i.e., we can access the rows randomly, i.e., we have pagination). Should we provide a way to increase or remove this limit?

closed

2024-06-03T08:55:08Z

2024-07-22T11:32:49Z

2024-07-11T15:04:04Z

severo

Update ruff to 0.4.7

Update ruff to 0.4.7.

Update ruff to 0.4.7: Update ruff to 0.4.7.

closed

2024-06-03T07:08:23Z

2024-06-03T08:46:25Z

2024-06-03T08:46:24Z

albertvillanova

Feature Request: Freeze/Restart/Log Viewer Option for Users.

### Description I've noticed a couple of things that could make using the dataset-viewer better. Here are three simple suggestions based on what I've experienced: ### Suggestions 1. **Keeping Parquet Viewer Steady** - Issue: Every time I tweak something like the content in the README, the Parquet viewer resets and tries to rebuild it. - Solution Idea: Add a "freeze" button for users so it stays where it is. 2. **Easier Viewer Restart** - Issue: It's definitely annoying to have to restart the viewer manually all the time for the dataset-viewer maintainers. - Solution Idea: Create a "restart" button for users so they can quickly reset the viewer. 3. **Log** - Issue: Just having access to the log of dataset-viewer errors would be very helpful. But the current log report is too abstract! - Solution Idea: Improve the error log report for better understanding. Thanks for considering these suggestions!

Feature Request: Freeze/Restart/Log Viewer Option for Users.: ### Description I've noticed a couple of things that could make using the dataset-viewer better. Here are three simple suggestions based on what I've experienced: ### Suggestions 1. **Keeping Parquet Viewer Steady** - Issue: Every time I tweak something like the content in the README, the Parquet viewer resets and tries to rebuild it. - Solution Idea: Add a "freeze" button for users so it stays where it is. 2. **Easier Viewer Restart** - Issue: It's definitely annoying to have to restart the viewer manually all the time for the dataset-viewer maintainers. - Solution Idea: Create a "restart" button for users so they can quickly reset the viewer. 3. **Log** - Issue: Just having access to the log of dataset-viewer errors would be very helpful. But the current log report is too abstract! - Solution Idea: Improve the error log report for better understanding. Thanks for considering these suggestions!

closed

2024-05-30T17:27:23Z

2024-05-31T08:51:34Z

2024-05-31T08:51:33Z

kargaranamir

Create a new error code (retryable) for "Consistency check failed"

See https://huggingface.co/datasets/cis-lmu/Taxi1500-RawData/viewer/mal_Mlym It currently gives this error: ``` Consistency check failed: file should be of size 13393376 but has size 11730856 ((…)a9b50f83/mal_Mlym/taxi1500/dataset.arrow). We are sorry for the inconvenience. Please retry download and pass `force_download=True, resume_download=False` as argument. If the issue persists, please let us know by opening an issue on https://github.com/huggingface/huggingface_hub. ``` discussion here: https://huggingface.co/datasets/cis-lmu/Taxi1500-RawData/discussions/1 We should catch the exception and raise a new error, that should be a "retryable" error.

Create a new error code (retryable) for "Consistency check failed": See https://huggingface.co/datasets/cis-lmu/Taxi1500-RawData/viewer/mal_Mlym It currently gives this error: ``` Consistency check failed: file should be of size 13393376 but has size 11730856 ((…)a9b50f83/mal_Mlym/taxi1500/dataset.arrow). We are sorry for the inconvenience. Please retry download and pass `force_download=True, resume_download=False` as argument. If the issue persists, please let us know by opening an issue on https://github.com/huggingface/huggingface_hub. ``` discussion here: https://huggingface.co/datasets/cis-lmu/Taxi1500-RawData/discussions/1 We should catch the exception and raise a new error, that should be a "retryable" error.

closed

2024-05-30T16:42:22Z

2024-05-31T08:18:47Z

2024-05-31T08:18:47Z

severo

Re-add torch dependency

needed for webdataset with .pth files in them

Re-add torch dependency: needed for webdataset with .pth files in them

closed

2024-05-30T14:22:00Z

2024-05-30T16:15:46Z

2024-05-30T16:15:45Z

lhoestq

add "duration" field to audio cells

As it was done with width/height for image cells: https://github.com/huggingface/dataset-viewer/pull/600 It will help to show additional information in the dataset viewer, in particular: highlighting a row will select the appropriate bar in the durations histogram.

add "duration" field to audio cells: As it was done with width/height for image cells: https://github.com/huggingface/dataset-viewer/pull/600 It will help to show additional information in the dataset viewer, in particular: highlighting a row will select the appropriate bar in the durations histogram.

open

2024-05-30T13:00:16Z

2024-05-30T16:26:52Z

severo

BFF endpoint to replace multiple parallel requests

WIP

BFF endpoint to replace multiple parallel requests: WIP

closed

2024-05-30T10:22:45Z

2024-07-29T11:37:24Z

2024-07-07T15:04:11Z

severo