id int64 959M 2.55B | title stringlengths 3 133 | body stringlengths 1 65.5k ⌀ | description stringlengths 5 65.6k | state stringclasses 2
values | created_at stringlengths 20 20 | updated_at stringlengths 20 20 | closed_at stringlengths 20 20 ⌀ | user stringclasses 174
values |
|---|---|---|---|---|---|---|---|---|
2,075,203,933 | reduce resources for now | null | reduce resources for now: | closed | 2024-01-10T21:05:24Z | 2024-01-10T21:08:33Z | 2024-01-10T21:08:32Z | severo |
2,074,892,546 | Update gitpython dependency to fix vulnerability | This should fix 12 dependabot alerts.
Note that all Poetry lock files are modified by this PR because a previous PR used a different version of Poetry (1.7.1 instead of 1.4.2): @lhoestq
- #2231
This PR reverts the lock files to the ones generated by Poetry version 1.4.2.
Fix #2268. | Update gitpython dependency to fix vulnerability: This should fix 12 dependabot alerts.
Note that all Poetry lock files are modified by this PR because a previous PR used a different version of Poetry (1.7.1 instead of 1.4.2): @lhoestq
- #2231
This PR reverts the lock files to the ones generated by Poetry vers... | closed | 2024-01-10T17:57:31Z | 2024-01-10T21:02:49Z | 2024-01-10T21:02:48Z | albertvillanova |
2,074,876,816 | hardcode the error codes to retry | There is no need to have this as an environment variable, and it clutters the code.
Note that the two lists we had (default one and backfill one) were different, by the way | hardcode the error codes to retry: There is no need to have this as an environment variable, and it clutters the code.
Note that the two lists we had (default one and backfill one) were different, by the way | closed | 2024-01-10T17:47:30Z | 2024-01-10T21:19:12Z | 2024-01-10T21:19:11Z | severo |
2,074,852,189 | Vulnerability in gitpython < 3.1.41 | Update gitpython dependency to fix vulnerability. | Vulnerability in gitpython < 3.1.41: Update gitpython dependency to fix vulnerability. | closed | 2024-01-10T17:31:52Z | 2024-01-10T21:02:49Z | 2024-01-10T21:02:49Z | albertvillanova |
2,074,810,169 | Bump gitpython from 3.1.40 to 3.1.41 in /docs | Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.40 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<h2>3.1.41 - fix Windows security issue</h2>
<p>... | Bump gitpython from 3.1.40 to 3.1.41 in /docs: Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.40 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<... | closed | 2024-01-10T17:09:38Z | 2024-01-10T21:03:37Z | 2024-01-10T21:03:35Z | dependabot[bot] |
2,074,789,185 | Bump gitpython from 3.1.37 to 3.1.41 in /services/worker | Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<h2>3.1.41 - fix Windows security issue</h2>
<p>... | Bump gitpython from 3.1.37 to 3.1.41 in /services/worker: Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<bl... | closed | 2024-01-10T16:58:43Z | 2024-01-10T21:09:28Z | 2024-01-10T21:09:19Z | dependabot[bot] |
2,074,786,526 | Bump gitpython from 3.1.37 to 3.1.41 in /services/search | Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<h2>3.1.41 - fix Windows security issue</h2>
<p>... | Bump gitpython from 3.1.37 to 3.1.41 in /services/search: Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<bl... | closed | 2024-01-10T16:57:13Z | 2024-01-10T21:09:23Z | 2024-01-10T21:09:12Z | dependabot[bot] |
2,074,786,165 | Bump gitpython from 3.1.37 to 3.1.41 in /libs/libcommon | Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<h2>3.1.41 - fix Windows security issue</h2>
<p>... | Bump gitpython from 3.1.37 to 3.1.41 in /libs/libcommon: Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blo... | closed | 2024-01-10T16:57:01Z | 2024-01-10T21:09:22Z | 2024-01-10T21:09:11Z | dependabot[bot] |
2,074,785,462 | Bump gitpython from 3.1.37 to 3.1.41 in /services/admin | Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<h2>3.1.41 - fix Windows security issue</h2>
<p>... | Bump gitpython from 3.1.37 to 3.1.41 in /services/admin: Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blo... | closed | 2024-01-10T16:56:37Z | 2024-01-10T21:09:21Z | 2024-01-10T21:09:11Z | dependabot[bot] |
2,074,785,347 | Bump gitpython from 3.1.37 to 3.1.41 in /jobs/mongodb_migration | Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<h2>3.1.41 - fix Windows security issue</h2>
<p>... | Bump gitpython from 3.1.37 to 3.1.41 in /jobs/mongodb_migration: Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em><... | closed | 2024-01-10T16:56:32Z | 2024-01-10T21:09:19Z | 2024-01-10T21:09:10Z | dependabot[bot] |
2,074,785,253 | Bump gitpython from 3.1.37 to 3.1.41 in /services/rows | Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<h2>3.1.41 - fix Windows security issue</h2>
<p>... | Bump gitpython from 3.1.37 to 3.1.41 in /services/rows: Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<bloc... | closed | 2024-01-10T16:56:29Z | 2024-01-10T21:09:18Z | 2024-01-10T21:09:09Z | dependabot[bot] |
2,074,785,131 | Bump gitpython from 3.1.37 to 3.1.41 in /jobs/cache_maintenance | Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<h2>3.1.41 - fix Windows security issue</h2>
<p>... | Bump gitpython from 3.1.37 to 3.1.41 in /jobs/cache_maintenance: Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em><... | closed | 2024-01-10T16:56:26Z | 2024-01-10T21:09:17Z | 2024-01-10T21:09:09Z | dependabot[bot] |
2,074,784,890 | Bump gitpython from 3.1.37 to 3.1.41 in /services/sse-api | Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<h2>3.1.41 - fix Windows security issue</h2>
<p>... | Bump gitpython from 3.1.37 to 3.1.41 in /services/sse-api: Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<b... | closed | 2024-01-10T16:56:18Z | 2024-01-10T21:09:16Z | 2024-01-10T21:09:08Z | dependabot[bot] |
2,074,783,121 | Bump gitpython from 3.1.37 to 3.1.41 in /e2e | Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<h2>3.1.41 - fix Windows security issue</h2>
<p>... | Bump gitpython from 3.1.37 to 3.1.41 in /e2e: Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<h... | closed | 2024-01-10T16:55:15Z | 2024-01-10T21:09:17Z | 2024-01-10T21:09:07Z | dependabot[bot] |
2,074,781,612 | Bump gitpython from 3.1.37 to 3.1.41 in /services/api | Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<h2>3.1.41 - fix Windows security issue</h2>
<p>... | Bump gitpython from 3.1.37 to 3.1.41 in /services/api: Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<block... | closed | 2024-01-10T16:54:23Z | 2024-01-10T21:09:15Z | 2024-01-10T21:09:07Z | dependabot[bot] |
2,074,774,124 | Bump gitpython from 3.1.37 to 3.1.41 in /libs/libapi | Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockquote>
<h2>3.1.41 - fix Windows security issue</h2>
<p>... | Bump gitpython from 3.1.37 to 3.1.41 in /libs/libapi: Bumps [gitpython](https://github.com/gitpython-developers/GitPython) from 3.1.37 to 3.1.41.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/gitpython-developers/GitPython/releases">gitpython's releases</a>.</em></p>
<blockq... | closed | 2024-01-10T16:50:15Z | 2024-01-10T21:09:16Z | 2024-01-10T21:09:06Z | dependabot[bot] |
2,074,730,786 | filter webhook scope | it was updating datasets completely on event scopes like "discussion.comment" | filter webhook scope: it was updating datasets completely on event scopes like "discussion.comment" | closed | 2024-01-10T16:26:44Z | 2024-01-10T18:01:02Z | 2024-01-10T18:01:01Z | lhoestq |
2,074,223,303 | Dont delete dataset cache if webhook is update on another branch | close https://github.com/huggingface/datasets-server/issues/2253 | Dont delete dataset cache if webhook is update on another branch: close https://github.com/huggingface/datasets-server/issues/2253 | closed | 2024-01-10T12:00:49Z | 2024-01-10T12:21:50Z | 2024-01-10T12:21:48Z | lhoestq |
2,073,920,913 | Cache entries seem to be deleted again and again if dataset page is accessed when the jobs are running | I observed on https://huggingface.co/datasets/wisenut-nlp-team/FiD-Merge-Dataset_temp_addlong:
- erroneous contents: https://huggingface.co/datasets/wisenut-nlp-team/FiD-Merge-Dataset_temp_addlong/discussions/2
- I recreated the dataset through the admin UI
- some cache entries were created, the config-parquet-and... | Cache entries seem to be deleted again and again if dataset page is accessed when the jobs are running: I observed on https://huggingface.co/datasets/wisenut-nlp-team/FiD-Merge-Dataset_temp_addlong:
- erroneous contents: https://huggingface.co/datasets/wisenut-nlp-team/FiD-Merge-Dataset_temp_addlong/discussions/2
-... | closed | 2024-01-10T09:13:14Z | 2024-01-10T12:21:49Z | 2024-01-10T12:21:49Z | severo |
2,072,763,726 | Bump fonttools from 4.39.4 to 4.43.0 in /front/admin_ui | Bumps [fonttools](https://github.com/fonttools/fonttools) from 4.39.4 to 4.43.0.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/fonttools/fonttools/releases">fonttools's releases</a>.</em></p>
<blockquote>
<h2>4.43.0</h2>
<ul>
<li>[subset] Set up lxml <code>XMLParser(resolve_... | Bump fonttools from 4.39.4 to 4.43.0 in /front/admin_ui: Bumps [fonttools](https://github.com/fonttools/fonttools) from 4.39.4 to 4.43.0.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/fonttools/fonttools/releases">fonttools's releases</a>.</em></p>
<blockquote>
<h2>4.43.0</h... | closed | 2024-01-09T16:48:24Z | 2024-01-09T17:13:06Z | 2024-01-09T17:13:05Z | dependabot[bot] |
2,072,515,738 | Finish the post-messages script earlier | The post-messages have been running for more than 15 hours now even if it does not take more than 2 minutes to complete.
https://cd.internal.huggingface.tech/applications/argo-cd/datasets-server-prod?view=pods&node=%2FPod%2Fdatasets-server%2Fprod-datasets-server-job-post-messages-28412650-w4x4c%2F0
<img width="89... | Finish the post-messages script earlier: The post-messages have been running for more than 15 hours now even if it does not take more than 2 minutes to complete.
https://cd.internal.huggingface.tech/applications/argo-cd/datasets-server-prod?view=pods&node=%2FPod%2Fdatasets-server%2Fprod-datasets-server-job-post-mess... | closed | 2024-01-09T14:43:47Z | 2024-01-22T14:10:11Z | 2024-01-22T14:10:11Z | severo |
2,072,511,934 | Run the backfill script in parallel? | The backfilll script is very slow (one dataset at a time) and also it breaks due to OOM (https://cd.internal.huggingface.tech/applications/argo-cd/datasets-server-prod?view=pods&node=%2FPod%2Fdatasets-server%2Fprod-datasets-server-job-backfill-28412480-2n9ps%2F0)
Maybe we could run the loop iterations in parallel, a... | Run the backfill script in parallel?: The backfilll script is very slow (one dataset at a time) and also it breaks due to OOM (https://cd.internal.huggingface.tech/applications/argo-cd/datasets-server-prod?view=pods&node=%2FPod%2Fdatasets-server%2Fprod-datasets-server-job-backfill-28412480-2n9ps%2F0)
Maybe we could ... | closed | 2024-01-09T14:41:51Z | 2024-04-05T09:45:38Z | 2024-04-05T09:45:38Z | severo |
2,072,493,963 | remove CACHE_MAX_DAYS feature | it simplifies the code, and it avoids periodic waste of resources to refresh entries that are already OK. If we want to refresh old entries, we can do it easily.
Requires https://github.com/huggingface/datasets-server/pull/2248 to be merged first
Closes https://github.com/huggingface/datasets-server/issues/1219 w... | remove CACHE_MAX_DAYS feature: it simplifies the code, and it avoids periodic waste of resources to refresh entries that are already OK. If we want to refresh old entries, we can do it easily.
Requires https://github.com/huggingface/datasets-server/pull/2248 to be merged first
Closes https://github.com/huggingfac... | closed | 2024-01-09T14:32:42Z | 2024-01-09T15:45:16Z | 2024-01-09T15:45:15Z | severo |
2,072,439,614 | recreate the dataset on webhook to avoid mixed content | fixes #1823. Note that we now always require the user to be trusted (ie to pass a secret with the /webhook request) because all the actions are destructive.
Depends on merging https://github.com/huggingface/datasets-server/pull/2246 first | recreate the dataset on webhook to avoid mixed content: fixes #1823. Note that we now always require the user to be trusted (ie to pass a secret with the /webhook request) because all the actions are destructive.
Depends on merging https://github.com/huggingface/datasets-server/pull/2246 first | closed | 2024-01-09T14:05:59Z | 2024-01-09T15:42:20Z | 2024-01-09T15:42:18Z | severo |
2,072,272,653 | Change "retries" for "failed_runs" in cache collection | For aggregated steps, the "retries" field is increased each time a child is processed, which can prevent retrying jobs when there is an error.
See for example:
This cache has been updated 11 times because o... | Change "retries" for "failed_runs" in cache collection: For aggregated steps, the "retries" field is increased each time a child is processed, which can prevent retrying jobs when there is an error.
See for example:
, saving the support for private datasets itself.
The most important part is that we now control a dataset only with:
- operations.update()
- operations.delete()
It's where we chec... | Refactor dataset updates but dont support private datasets: This PR contains the refactoring needed for supporting the private datasets (https://github.com/huggingface/datasets-server/pull/2224), saving the support for private datasets itself.
The most important part is that we now control a dataset only with:
- op... | closed | 2024-01-08T14:36:08Z | 2024-01-08T22:25:03Z | 2024-01-08T19:56:35Z | severo |
2,070,491,993 | allow speechcolab/gigaspeech | null | allow speechcolab/gigaspeech: | closed | 2024-01-08T13:44:20Z | 2024-01-08T14:24:32Z | 2024-01-08T14:24:31Z | lhoestq |
2,070,402,697 | add filter column + ignore errors (fixes #2241) | fixes #2241 | add filter column + ignore errors (fixes #2241): fixes #2241 | closed | 2024-01-08T12:55:06Z | 2024-01-09T13:19:06Z | 2024-01-08T13:09:53Z | severo |
2,070,394,348 | The admin UI is not able to handle some gated datasets | See
<img width="1551" alt="Capture d’écran 2024-01-08 à 13 47 41" src="https://github.com/huggingface/datasets-server/assets/1676121/daa08d5e-a1b1-4416-adf8-ae69573a2daa">
<img width="1034" alt="Capture d’écran 2024-01-08 à 13 48 27" src="https://github.com/huggingface/datasets-server/assets/1676121/9154e421-e53... | The admin UI is not able to handle some gated datasets: See
<img width="1551" alt="Capture d’écran 2024-01-08 à 13 47 41" src="https://github.com/huggingface/datasets-server/assets/1676121/daa08d5e-a1b1-4416-adf8-ae69573a2daa">
<img width="1034" alt="Capture d’écran 2024-01-08 à 13 48 27" src="https://github.com... | closed | 2024-01-08T12:50:08Z | 2024-01-08T13:09:54Z | 2024-01-08T13:09:54Z | severo |
2,070,125,988 | add .ogg as supported audio format | null | add .ogg as supported audio format: | closed | 2024-01-08T10:14:07Z | 2024-01-08T10:17:15Z | 2024-01-08T10:14:20Z | severo |
2,070,119,009 | add missing 'filter' field in IsValidResponse | null | add missing 'filter' field in IsValidResponse: | closed | 2024-01-08T10:09:53Z | 2024-01-08T10:13:10Z | 2024-01-08T10:10:08Z | severo |
2,067,798,964 | add JobManagerCrashedError to errors list | null | add JobManagerCrashedError to errors list: | closed | 2024-01-05T17:56:34Z | 2024-01-05T18:07:54Z | 2024-01-05T18:07:53Z | AndreaFrancis |
2,067,789,128 | Bump pycryptodomex from 3.18.0 to 3.19.1 in /services/worker | Bumps [pycryptodomex](https://github.com/Legrandin/pycryptodome) from 3.18.0 to 3.19.1.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/Legrandin/pycryptodome/releases">pycryptodomex's releases</a>.</em></p>
<blockquote>
<h2>v3.19.1 - Zeil</h2>
<h2>Resolved issues</h2>
<ul>
<l... | Bump pycryptodomex from 3.18.0 to 3.19.1 in /services/worker: Bumps [pycryptodomex](https://github.com/Legrandin/pycryptodome) from 3.18.0 to 3.19.1.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/Legrandin/pycryptodome/releases">pycryptodomex's releases</a>.</em></p>
<blockq... | closed | 2024-01-05T17:49:46Z | 2024-01-08T07:14:05Z | 2024-01-08T07:14:04Z | dependabot[bot] |
2,067,221,090 | add filter field to is-valid | fixes #2235 | add filter field to is-valid: fixes #2235 | closed | 2024-01-05T11:58:19Z | 2024-01-05T18:21:34Z | 2024-01-05T18:21:33Z | severo |
2,067,195,944 | Add field "filter" to /is-valid endpoint | Currently, in moon-landing, we use the "search" field to enable the search and the filter. But on https://huggingface.co/datasets/Andyrasika/ShoeSandalBootimages/viewer/default/train?p=1, for example, the search is disabled because we have no text columns, while the filter works (see https://datasets-server.huggingface... | Add field "filter" to /is-valid endpoint: Currently, in moon-landing, we use the "search" field to enable the search and the filter. But on https://huggingface.co/datasets/Andyrasika/ShoeSandalBootimages/viewer/default/train?p=1, for example, the search is disabled because we have no text columns, while the filter work... | closed | 2024-01-05T11:37:53Z | 2024-01-05T18:21:34Z | 2024-01-05T18:21:34Z | severo |
2,065,609,452 | chore: 🤖 use the same packages in all the docker imges | we could surely build a base image, if somebody wants to do it...
it should fix the following error message we see in the admin service logs:
```
/src/services/admin/.venv/lib/python3.9/site-packages/pydub/utils.py:170: RuntimeWarning: Couldn't find ffmpeg or avconv - defaulting to ffmpeg, but may not work
w... | chore: 🤖 use the same packages in all the docker imges: we could surely build a base image, if somebody wants to do it...
it should fix the following error message we see in the admin service logs:
```
/src/services/admin/.venv/lib/python3.9/site-packages/pydub/utils.py:170: RuntimeWarning: Couldn't find ffmpeg... | closed | 2024-01-04T12:52:11Z | 2024-01-04T15:14:28Z | 2024-01-04T15:14:27Z | severo |
2,064,567,746 | change cookie to fix e2e tests | null | change cookie to fix e2e tests: | closed | 2024-01-03T19:53:11Z | 2024-01-03T22:27:11Z | 2024-01-03T22:27:10Z | severo |
2,064,211,063 | Set 3.9.18 for pyenv users | null | Set 3.9.18 for pyenv users: | closed | 2024-01-03T15:19:45Z | 2024-01-03T15:30:15Z | 2024-01-03T15:30:14Z | lhoestq |
2,064,187,804 | Update `datasets` to 2.16.1 | ...and hfh to 0.20.1
This enables:
- Webdataset support for the Dataset Viewer
- Faster data files resolution, which makes listing configs, splits and generating parquet export faster
- Less pressure on the HF back-end (less /tree calls on the Hub, and they no longer use expand=true)
In addition to the updat... | Update `datasets` to 2.16.1: ...and hfh to 0.20.1
This enables:
- Webdataset support for the Dataset Viewer
- Faster data files resolution, which makes listing configs, splits and generating parquet export faster
- Less pressure on the HF back-end (less /tree calls on the Hub, and they no longer use expand=true... | closed | 2024-01-03T15:04:28Z | 2024-01-05T15:34:55Z | 2024-01-05T09:57:12Z | lhoestq |
2,063,636,123 | Group endpoints for calls from the Hub (back for front) | In [moonlanding](https://github.com/huggingface/moon-landing/pull/8565#discussion_r1440224297) (internal) we call several endpoints in parallel. We could group them in one call to datasets-server, with all the available information in one response. | Group endpoints for calls from the Hub (back for front): In [moonlanding](https://github.com/huggingface/moon-landing/pull/8565#discussion_r1440224297) (internal) we call several endpoints in parallel. We could group them in one call to datasets-server, with all the available information in one response. | open | 2024-01-03T09:46:55Z | 2024-08-02T08:54:05Z | null | severo |
2,063,029,370 | Upgrade to datasets@2.16.1 | We're still using 2.15.0 from November! We could upgrade to 2.16.1. Release notes: https://github.com/huggingface/datasets/releases/tag/2.16.0 and https://github.com/huggingface/datasets/releases/tag/2.16.1
Main changes to to in datasets-server:
- remove custom code to disable datasets script, and simply pass `trus... | Upgrade to datasets@2.16.1: We're still using 2.15.0 from November! We could upgrade to 2.16.1. Release notes: https://github.com/huggingface/datasets/releases/tag/2.16.0 and https://github.com/huggingface/datasets/releases/tag/2.16.1
Main changes to to in datasets-server:
- remove custom code to disable datasets s... | closed | 2024-01-02T22:31:51Z | 2024-01-05T09:57:13Z | 2024-01-05T09:57:13Z | severo |
2,062,757,000 | The admin UI shows an error | <img width="1050" alt="Capture d’écran 2024-01-02 à 18 55 20" src="https://github.com/huggingface/datasets-server/assets/1676121/2a24e7ca-1cda-42e0-b833-33688c2baf21">
| The admin UI shows an error: <img width="1050" alt="Capture d’écran 2024-01-02 à 18 55 20" src="https://github.com/huggingface/datasets-server/assets/1676121/2a24e7ca-1cda-42e0-b833-33688c2baf21">
| closed | 2024-01-02T17:55:34Z | 2024-01-18T14:57:00Z | 2024-01-18T14:57:00Z | severo |
2,060,246,830 | Fix stats for datasets bigger than 5Gb on first n parquets | Currently for datasets with lots of parquet files which were copied from the main branch statistics are not computed with `SplitWithTooBigParquetError` (example: `bigcode/starcoderdata`). While we can take first n parquet files instead, as it's done for `duckdb-index` runner.
\+ also updated openapi.json with a cha... | Fix stats for datasets bigger than 5Gb on first n parquets: Currently for datasets with lots of parquet files which were copied from the main branch statistics are not computed with `SplitWithTooBigParquetError` (example: `bigcode/starcoderdata`). While we can take first n parquet files instead, as it's done for `duckd... | closed | 2023-12-29T16:39:48Z | 2024-02-15T21:17:57Z | 2024-02-15T21:17:56Z | polinaeterna |
2,053,470,126 | Update gradio dependency to fix vulnerability | Update gradio to 4.11.0.
This should fix 2 dependabot alerts. | Update gradio dependency to fix vulnerability: Update gradio to 4.11.0.
This should fix 2 dependabot alerts. | closed | 2023-12-22T06:51:08Z | 2023-12-22T06:53:54Z | 2023-12-22T06:53:53Z | albertvillanova |
2,052,904,265 | Bump gradio from 3.34.0 to 4.11.0 in /front/admin_ui | Bumps [gradio](https://github.com/gradio-app/gradio) from 3.34.0 to 4.11.0.
<details>
<summary>Changelog</summary>
<p><em>Sourced from <a href="https://github.com/gradio-app/gradio/blob/main/CHANGELOG.md">gradio's changelog</a>.</em></p>
<blockquote>
<h2>4.11.0</h2>
<h3>Features</h3>
<ul>
<li><a href="https://redirect.... | Bump gradio from 3.34.0 to 4.11.0 in /front/admin_ui: Bumps [gradio](https://github.com/gradio-app/gradio) from 3.34.0 to 4.11.0.
<details>
<summary>Changelog</summary>
<p><em>Sourced from <a href="https://github.com/gradio-app/gradio/blob/main/CHANGELOG.md">gradio's changelog</a>.</em></p>
<blockquote>
<h2>4.11.0</h2>... | closed | 2023-12-21T18:40:22Z | 2023-12-22T06:58:55Z | 2023-12-22T06:57:52Z | dependabot[bot] |
2,052,821,051 | Support private datasets | This PR changes the logic a bit: operations.py replaces dataset.py and is the module that must be used to update or delete a dataset (ie: do not create jobs or cache entries directly or through DatasetOrchestrator). Indeed, operations.py checks if the dataset is supported (public, or private + pro user or enterprise or... | Support private datasets: This PR changes the logic a bit: operations.py replaces dataset.py and is the module that must be used to update or delete a dataset (ie: do not create jobs or cache entries directly or through DatasetOrchestrator). Indeed, operations.py checks if the dataset is supported (public, or private +... | closed | 2023-12-21T17:31:28Z | 2024-01-30T21:09:26Z | 2024-01-30T21:09:25Z | severo |
2,052,234,690 | update transformers in services/workers | For https://github.com/huggingface/datasets-server/security/dependabot/347 | update transformers in services/workers: For https://github.com/huggingface/datasets-server/security/dependabot/347 | closed | 2023-12-21T11:21:07Z | 2023-12-21T11:32:27Z | 2023-12-21T11:32:26Z | AndreaFrancis |
2,051,273,738 | Bump transformers from 4.30.0 to 4.36.0 in /services/worker | Bumps [transformers](https://github.com/huggingface/transformers) from 4.30.0 to 4.36.0.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/huggingface/transformers/releases">transformers's releases</a>.</em></p>
<blockquote>
<h2>v4.36: Mixtral, Llava/BakLlava, SeamlessM4T v2, AM... | Bump transformers from 4.30.0 to 4.36.0 in /services/worker: Bumps [transformers](https://github.com/huggingface/transformers) from 4.30.0 to 4.36.0.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/huggingface/transformers/releases">transformers's releases</a>.</em></p>
<block... | closed | 2023-12-20T21:11:46Z | 2023-12-21T11:33:08Z | 2023-12-21T11:33:07Z | dependabot[bot] |
2,050,811,325 | Retry JobManagerCrashedError and limit a certain number of retries | Proposal for https://github.com/huggingface/datasets-server/issues/2218
When saving a new cache result, the "retries" counter will be increased if there already exists another record for the same revision (and kind+dataset+config+split combination).
If it is the first time it is by default 0.
If there already exi... | Retry JobManagerCrashedError and limit a certain number of retries: Proposal for https://github.com/huggingface/datasets-server/issues/2218
When saving a new cache result, the "retries" counter will be increased if there already exists another record for the same revision (and kind+dataset+config+split combination).... | closed | 2023-12-20T15:36:16Z | 2024-01-05T17:42:15Z | 2024-01-05T16:37:18Z | AndreaFrancis |
2,050,775,746 | Remove abstractions | - import shared instance of processing_graph from libcommon.processing_graph, instead of instanciating a new one from a (fake) configuration, and passing it through function arguments.
- fix tests locally (`/storage` directory does not exist on every machine; and should not be used for tests) | Remove abstractions: - import shared instance of processing_graph from libcommon.processing_graph, instead of instanciating a new one from a (fake) configuration, and passing it through function arguments.
- fix tests locally (`/storage` directory does not exist on every machine; and should not be used for tests) | closed | 2023-12-20T15:24:10Z | 2023-12-20T16:33:58Z | 2023-12-20T16:33:57Z | severo |
2,048,892,846 | Remove abstractions | Hardcode step names, instead of creating an abstraction around what a step provides (it's too artificial, and we only have one DAG, so: no need to look for genericity) | Remove abstractions: Hardcode step names, instead of creating an abstraction around what a step provides (it's too artificial, and we only have one DAG, so: no need to look for genericity) | closed | 2023-12-19T15:28:12Z | 2023-12-20T13:08:39Z | 2023-12-20T13:08:38Z | severo |
2,048,882,493 | JobManagerCrashedError jobs are never retried | Currently, we have 7768 jobs with error_code `JobManagerCrashedError`. Some of them are caused by zombie killer set crashes.
```
Atlas atlas-x5jgb3-shard-0 [primary] datasets_server_cache> db.cachedResponsesBlue.aggregate([{$match:{error_code:"JobManagerCrashedError","details.copied_from_artifact":{$exists:false}}}... | JobManagerCrashedError jobs are never retried: Currently, we have 7768 jobs with error_code `JobManagerCrashedError`. Some of them are caused by zombie killer set crashes.
```
Atlas atlas-x5jgb3-shard-0 [primary] datasets_server_cache> db.cachedResponsesBlue.aggregate([{$match:{error_code:"JobManagerCrashedError","... | closed | 2023-12-19T15:22:30Z | 2024-01-09T20:32:58Z | 2024-01-09T20:32:57Z | AndreaFrancis |
2,048,748,961 | refactor: 💡 remove useless processing_step parameter | it's ridiculous... it was never used!!! | refactor: 💡 remove useless processing_step parameter: it's ridiculous... it was never used!!! | closed | 2023-12-19T14:15:21Z | 2023-12-19T15:06:38Z | 2023-12-19T15:06:37Z | severo |
2,048,695,794 | refactor: 💡 remove useless get_job_runner_version() | null | refactor: 💡 remove useless get_job_runner_version(): | closed | 2023-12-19T13:46:35Z | 2023-12-19T14:40:59Z | 2023-12-19T14:40:58Z | severo |
2,046,635,866 | Ensure /first-rows can always be created when truncating (ie: avoid TooBigContentError) | See https://huggingface.co/datasets/mikehemberger/inat_2021_train_mini_plantae
```
The size of the content of the first rows (358268 B) exceeds the maximum supported size (200000 B) even after truncation. Please report the issue.
Error code: TooBigContentError
```
Initially reported in https://github.com/h... | Ensure /first-rows can always be created when truncating (ie: avoid TooBigContentError) : See https://huggingface.co/datasets/mikehemberger/inat_2021_train_mini_plantae
```
The size of the content of the first rows (358268 B) exceeds the maximum supported size (200000 B) even after truncation. Please report the iss... | open | 2023-12-18T13:10:53Z | 2024-08-09T19:45:40Z | null | severo |
2,043,808,570 | increase limits memory for heavy workers | With 34Gi it fails because of OOMs when processing `split-duckdb-index` for some Wikipedia configs like 20231101.es and 20231101.en when creating index step.
| increase limits memory for heavy workers: With 34Gi it fails because of OOMs when processing `split-duckdb-index` for some Wikipedia configs like 20231101.es and 20231101.en when creating index step.
| closed | 2023-12-15T14:15:59Z | 2023-12-18T12:23:25Z | 2023-12-15T14:37:11Z | AndreaFrancis |
2,043,623,788 | remove duckdb setting | Rollback of setting in https://github.com/huggingface/datasets-server/pull/2212. Heavy workers are taking more time when inserting.
| remove duckdb setting: Rollback of setting in https://github.com/huggingface/datasets-server/pull/2212. Heavy workers are taking more time when inserting.
| closed | 2023-12-15T12:14:08Z | 2023-12-15T12:28:29Z | 2023-12-15T12:28:28Z | AndreaFrancis |
2,042,563,908 | try preserve_insertion_order | Memory setting didn't work for big datasets. It took more time to process, rolling back and trying another configuration as recommended in https://duckdb.org/docs/guides/performance/how-to-tune-workloads#the-preserve_insertion_order-option | try preserve_insertion_order: Memory setting didn't work for big datasets. It took more time to process, rolling back and trying another configuration as recommended in https://duckdb.org/docs/guides/performance/how-to-tune-workloads#the-preserve_insertion_order-option | closed | 2023-12-14T21:54:45Z | 2023-12-14T22:03:21Z | 2023-12-14T22:03:20Z | AndreaFrancis |
2,042,417,919 | try duckdb resources configuration | `split-duckdb-index` fails mostly for big datasets in the create index step (PRAGMA create_fts_index) because of OOM.
Same as in https://github.com/huggingface/datasets-server/pull/2081/files, I would like to try custom duckdb configurations to see if that helps.
Also I noticed the close condition was never closed wh... | try duckdb resources configuration: `split-duckdb-index` fails mostly for big datasets in the create index step (PRAGMA create_fts_index) because of OOM.
Same as in https://github.com/huggingface/datasets-server/pull/2081/files, I would like to try custom duckdb configurations to see if that helps.
Also I noticed the... | closed | 2023-12-14T20:21:21Z | 2023-12-14T20:41:00Z | 2023-12-14T20:40:59Z | AndreaFrancis |
2,042,092,167 | Deprecate/remove MetricsMongoResource and METRICS_MONGOENGINE_ALIAS from code | It's only used in the migrations, and nobody will ever run them again from an old version. | Deprecate/remove MetricsMongoResource and METRICS_MONGOENGINE_ALIAS from code: It's only used in the migrations, and nobody will ever run them again from an old version. | closed | 2023-12-14T16:56:19Z | 2024-02-23T11:24:01Z | 2024-02-23T11:24:01Z | severo |
2,041,786,098 | duckdb-index log info for monitoring | null | duckdb-index log info for monitoring: | closed | 2023-12-14T14:12:09Z | 2023-12-14T14:21:03Z | 2023-12-14T14:21:01Z | AndreaFrancis |
2,041,761,992 | Add a collection with datasets infos | While working on enabling private datasets (#39) under conditions (isPro, isEnterprise), I thought we missed a place where we control the access to the dataset.
I think the first step in the DAG, instead of dataset-config-names, should be more about the dataset characteristics: if it's private or public, maybe if it... | Add a collection with datasets infos: While working on enabling private datasets (#39) under conditions (isPro, isEnterprise), I thought we missed a place where we control the access to the dataset.
I think the first step in the DAG, instead of dataset-config-names, should be more about the dataset characteristics: ... | closed | 2023-12-14T13:59:42Z | 2024-01-11T14:30:03Z | 2024-01-11T14:30:02Z | severo |
2,041,655,766 | Backfill job processes datasets with disabled viewer? | If I read the code correctly, the backfill cronjob does not check if the dataset viewer is disabled (`viewer: false` in the README).
If we want to implement the dataset viewer for private datasets, under conditions (isPro, isEnterprise), we will have to check these conditions before adding jobs. | Backfill job processes datasets with disabled viewer?: If I read the code correctly, the backfill cronjob does not check if the dataset viewer is disabled (`viewer: false` in the README).
If we want to implement the dataset viewer for private datasets, under conditions (isPro, isEnterprise), we will have to check th... | closed | 2023-12-14T13:01:53Z | 2024-02-06T16:03:10Z | 2024-02-06T16:03:10Z | severo |
2,041,348,705 | Revert "feat: 🎸 block big dataset (#2204)" | This reverts commit 87b93e8f246bc5af61c56808c48ea2fa4834b20d. | Revert "feat: 🎸 block big dataset (#2204)": This reverts commit 87b93e8f246bc5af61c56808c48ea2fa4834b20d. | closed | 2023-12-14T09:59:38Z | 2023-12-14T10:00:32Z | 2023-12-14T09:59:44Z | severo |
2,041,284,341 | feat: 🎸 block big dataset | It generates >1K jobs with difficulty 90, which will never be processed (too few "heavy" workers). Also note that the dataset is gated with manual approval, which limits the audience of dataset viewer / parquet conversion. | feat: 🎸 block big dataset: It generates >1K jobs with difficulty 90, which will never be processed (too few "heavy" workers). Also note that the dataset is gated with manual approval, which limits the audience of dataset viewer / parquet conversion. | closed | 2023-12-14T09:23:38Z | 2023-12-14T09:24:08Z | 2023-12-14T09:23:54Z | severo |
2,041,279,943 | chore: 🤖 update vscode settings | the value cannot be boolean in new versions of vscode. See https://stackoverflow.com/a/77637765/7351594 | chore: 🤖 update vscode settings: the value cannot be boolean in new versions of vscode. See https://stackoverflow.com/a/77637765/7351594 | closed | 2023-12-14T09:21:06Z | 2023-12-14T09:21:12Z | 2023-12-14T09:21:11Z | severo |
2,039,217,749 | decrease number of workers | the queue is empty | decrease number of workers: the queue is empty | closed | 2023-12-13T08:38:30Z | 2023-12-13T08:39:00Z | 2023-12-13T08:38:35Z | severo |
2,038,529,618 | temporary blocking atom-in-the-universe | null | temporary blocking atom-in-the-universe: | closed | 2023-12-12T20:49:56Z | 2023-12-12T20:50:34Z | 2023-12-12T20:50:02Z | AndreaFrancis |
2,037,904,528 | Add TTL to git branch locks | Fix https://github.com/huggingface/datasets-server/issues/2123
TODO:
- [x] fix tests | Add TTL to git branch locks: Fix https://github.com/huggingface/datasets-server/issues/2123
TODO:
- [x] fix tests | closed | 2023-12-12T14:43:16Z | 2023-12-13T10:24:18Z | 2023-12-13T10:24:17Z | lhoestq |
2,037,740,057 | Pin a specific commit when allowing a dataset script | We manually allow the dataset script on the following datasets:
https://github.com/huggingface/datasets-server/blob/036af6b704827e8f55bdeb2a250726cf2255686b/chart/values.yaml#L84-L87
We could add the option to pin them to a specific commit. Further commits that might introduce a security issue will therefore not ... | Pin a specific commit when allowing a dataset script: We manually allow the dataset script on the following datasets:
https://github.com/huggingface/datasets-server/blob/036af6b704827e8f55bdeb2a250726cf2255686b/chart/values.yaml#L84-L87
We could add the option to pin them to a specific commit. Further commits tha... | closed | 2023-12-12T13:19:28Z | 2023-12-12T13:26:53Z | 2023-12-12T13:26:53Z | severo |
2,037,511,361 | 30 -> 50 medium workers | null | 30 -> 50 medium workers: | closed | 2023-12-12T11:03:13Z | 2023-12-12T11:03:46Z | 2023-12-12T11:03:19Z | severo |
2,037,416,879 | allow script on google/fleurs | null | allow script on google/fleurs: | closed | 2023-12-12T10:10:41Z | 2023-12-12T10:11:13Z | 2023-12-12T10:10:46Z | severo |
2,035,746,803 | fix delete_jobs key in admin-ui | Fixes https://github.com/huggingface/datasets-server/issues/2195 | fix delete_jobs key in admin-ui: Fixes https://github.com/huggingface/datasets-server/issues/2195 | closed | 2023-12-11T13:59:25Z | 2023-12-11T13:59:43Z | 2023-12-11T13:59:32Z | AndreaFrancis |
2,035,680,419 | The "Recreate" admin tab does not work | <img width="1099" alt="Capture d’écran 2023-12-11 à 14 26 29" src="https://github.com/huggingface/datasets-server/assets/1676121/db50d616-5f3c-40c0-8df6-417c46277bf5">
| The "Recreate" admin tab does not work: <img width="1099" alt="Capture d’écran 2023-12-11 à 14 26 29" src="https://github.com/huggingface/datasets-server/assets/1676121/db50d616-5f3c-40c0-8df6-417c46277bf5">
| closed | 2023-12-11T13:27:05Z | 2023-12-11T14:05:37Z | 2023-12-11T13:59:50Z | severo |
2,031,453,974 | fix heartbeat error log | null | fix heartbeat error log: | closed | 2023-12-07T20:16:26Z | 2023-12-07T20:18:16Z | 2023-12-07T20:18:15Z | AndreaFrancis |
2,031,340,593 | Allow common voice | null | Allow common voice: | closed | 2023-12-07T18:50:57Z | 2023-12-07T18:51:35Z | 2023-12-07T18:51:22Z | lhoestq |
2,031,268,429 | logs and stop executor on failure | Sometimes, a long job gets finished by another worker (still investigating the cause), and heartbeat logs the issue forever.
Once the heartbeat fails for some reason, the executor should stop to avoid blocking the worker:
, and heartbeat logs the issue forever.
Once the heartbeat fails for some reason, the executor should stop to avoid blocking the worker:
![image](https://github.com/huggingface/datasets-server/asset... | closed | 2023-12-07T17:59:32Z | 2023-12-07T19:33:26Z | 2023-12-07T19:33:25Z | AndreaFrancis |
2,030,791,655 | More rows workers | `/rows` workers get killed since this morning when someone started using it to scrape coyo-700m
<img width="608" alt="image" src="https://github.com/huggingface/datasets-server/assets/42851186/53d960bc-3aed-48f5-85cf-a8725916ed1e">
<img width="685" alt="image" src="https://github.com/huggingface/datasets-server/a... | More rows workers: `/rows` workers get killed since this morning when someone started using it to scrape coyo-700m
<img width="608" alt="image" src="https://github.com/huggingface/datasets-server/assets/42851186/53d960bc-3aed-48f5-85cf-a8725916ed1e">
<img width="685" alt="image" src="https://github.com/huggingfac... | closed | 2023-12-07T13:43:10Z | 2023-12-12T11:12:21Z | 2023-12-12T11:12:21Z | lhoestq |
2,029,171,668 | rollback worker difficulty to previous values | null | rollback worker difficulty to previous values: | closed | 2023-12-06T18:40:52Z | 2023-12-06T18:44:16Z | 2023-12-06T18:44:15Z | AndreaFrancis |
2,028,990,172 | remove read_parquet from descriptive stats | Related to https://github.com/huggingface/datasets-server/issues/2047
I am not sure if this will help, but I saw that many jobs got stuck in the CREATE command, `duckdb-index` worker also performs this operation, but it looks like it doesnt stuck for so long.
I want to try loading directly to parquet files instead of... | remove read_parquet from descriptive stats: Related to https://github.com/huggingface/datasets-server/issues/2047
I am not sure if this will help, but I saw that many jobs got stuck in the CREATE command, `duckdb-index` worker also performs this operation, but it looks like it doesnt stuck for so long.
I want to try ... | closed | 2023-12-06T16:55:13Z | 2023-12-06T18:17:31Z | 2023-12-06T18:17:30Z | AndreaFrancis |
2,028,913,846 | removing useless status filter | Now that the queue only stores the status "started" and "waiting," it is not necessary to add filter `status__in=[Status.WAITING, Status.STARTED]`.
After this PR, I will monitor database performance to see if any index can be optimized/changed. | removing useless status filter: Now that the queue only stores the status "started" and "waiting," it is not necessary to add filter `status__in=[Status.WAITING, Status.STARTED]`.
After this PR, I will monitor database performance to see if any index can be optimized/changed. | closed | 2023-12-06T16:13:52Z | 2023-12-06T17:01:29Z | 2023-12-06T17:01:28Z | AndreaFrancis |
2,028,609,775 | Infer audio file extension | Infer audio file extension for audio filenames without extension (audio files downloaded from the Hub).
This will avoid unnecessary conversions of audio files without extension.
Follow-up of:
- #2180
Fix #2181. | Infer audio file extension: Infer audio file extension for audio filenames without extension (audio files downloaded from the Hub).
This will avoid unnecessary conversions of audio files without extension.
Follow-up of:
- #2180
Fix #2181. | closed | 2023-12-06T13:52:06Z | 2023-12-07T12:53:45Z | 2023-12-07T12:53:44Z | albertvillanova |
2,028,596,975 | temporary remove validation in migration | null | temporary remove validation in migration: | closed | 2023-12-06T13:45:36Z | 2023-12-06T13:51:13Z | 2023-12-06T13:51:12Z | AndreaFrancis |
2,028,344,081 | allow RedPajama-Data-1T | null | allow RedPajama-Data-1T: | closed | 2023-12-06T11:35:00Z | 2023-12-06T11:35:52Z | 2023-12-06T11:35:51Z | lhoestq |
2,027,026,389 | blocking temporary lunaluan datasets | Temporary blocking namespace lunaluan/* because of a high activity in datasets (many frequent commits):
- lunaluan/chatbox6_history 23778 commits
- lunaluan/chatbox8_history 22964 commits
- lunaluan/chatbox4_history 79255 commits
- lunaluan/chat_history5 46412 commits
- lunaluan/chatbox3_history 76140 commits | blocking temporary lunaluan datasets: Temporary blocking namespace lunaluan/* because of a high activity in datasets (many frequent commits):
- lunaluan/chatbox6_history 23778 commits
- lunaluan/chatbox8_history 22964 commits
- lunaluan/chatbox4_history 79255 commits
- lunaluan/chat_history5 46412 commits
- lun... | closed | 2023-12-05T20:03:02Z | 2023-12-05T20:05:47Z | 2023-12-05T20:05:46Z | AndreaFrancis |
2,026,996,460 | temporary changing difficulty intervals in workers | Currently we have a queue that isn't able to reduce for at least a couple of days
The current difficulty distribution is:
```
Atlas atlas-x5jgb3-shard-0 [primary] datasets_server_queue> db.jobsBlue.aggrega... | temporary changing difficulty intervals in workers: Currently we have a queue that isn't able to reduce for at least a couple of days
The current difficulty distribution is:
```
Atlas atlas-x5jgb3-shard-0 ... | closed | 2023-12-05T19:44:58Z | 2023-12-05T19:56:05Z | 2023-12-05T19:56:04Z | AndreaFrancis |
2,026,216,458 | use `polars` to compute stats over all datasets | \+ added unit-tests for each feature type
\+ added stats for boolean type (~~TODO: document + e2e test~~ - done) | use `polars` to compute stats over all datasets: \+ added unit-tests for each feature type
\+ added stats for boolean type (~~TODO: document + e2e test~~ - done) | closed | 2023-12-05T13:36:17Z | 2023-12-12T17:43:16Z | 2023-12-12T17:42:23Z | polinaeterna |
2,025,547,154 | Avoid unnecessary conversions of audio files without extension | This will improve the speed performance of the /rows endpoint.
As we analogously do in the `datasets` library and suggested by @lhoestq, we could infer the file format from the header of the file (magic bytes).
To be done, once this PR is merged:
- #2180 | Avoid unnecessary conversions of audio files without extension: This will improve the speed performance of the /rows endpoint.
As we analogously do in the `datasets` library and suggested by @lhoestq, we could infer the file format from the header of the file (magic bytes).
To be done, once this PR is merged:
- ... | closed | 2023-12-05T08:10:14Z | 2023-12-07T12:53:46Z | 2023-12-07T12:53:46Z | albertvillanova |
2,024,237,310 | Do not raise error if audio file has no extension | This PR enables support for audio no-script datasets on the Hub: audio files downloaded from the Hub have no extension, e.g. the audio column embedded in the Parquet file contains:
```
{'bytes': b'RIFF\x86Y\x0c\x00WAVEfmt \x10...',
'path': '1b607f504e9c4489e199198847691e2b6b60d99ef50614becb059ac71c22d737'}
```
... | Do not raise error if audio file has no extension: This PR enables support for audio no-script datasets on the Hub: audio files downloaded from the Hub have no extension, e.g. the audio column embedded in the Parquet file contains:
```
{'bytes': b'RIFF\x86Y\x0c\x00WAVEfmt \x10...',
'path': '1b607f504e9c4489e1991988... | closed | 2023-12-04T16:20:34Z | 2023-12-05T13:07:28Z | 2023-12-05T13:07:26Z | albertvillanova |
2,023,653,481 | remove finished_at migrations | As per the comment https://github.com/huggingface/datasets-server/pull/2175#discussion_r1412342252, remove finishet_at migrations will be handled in a separate PR.
Depends on https://github.com/huggingface/datasets-server/pull/2178
| remove finished_at migrations: As per the comment https://github.com/huggingface/datasets-server/pull/2175#discussion_r1412342252, remove finishet_at migrations will be handled in a separate PR.
Depends on https://github.com/huggingface/datasets-server/pull/2178
| closed | 2023-12-04T11:38:53Z | 2023-12-06T13:29:47Z | 2023-12-06T13:29:46Z | AndreaFrancis |
2,023,624,505 | remove finished_at from code | As suggested by @severo in https://github.com/huggingface/datasets-server/pull/2175#discussion_r1412342252 for "remove finished jobs," PR, finished_at, and TTL index field is being removed from code. Once merged, It will be necessary to deploy before changes in https://github.com/huggingface/datasets-server/pull/2175 | remove finished_at from code: As suggested by @severo in https://github.com/huggingface/datasets-server/pull/2175#discussion_r1412342252 for "remove finished jobs," PR, finished_at, and TTL index field is being removed from code. Once merged, It will be necessary to deploy before changes in https://github.com/huggingf... | closed | 2023-12-04T11:22:16Z | 2023-12-06T11:31:09Z | 2023-12-06T11:31:08Z | AndreaFrancis |
2,020,973,967 | Add filter endpoint to Quickstart docs | Add filter endpoint to Quickstart docs. | Add filter endpoint to Quickstart docs: Add filter endpoint to Quickstart docs. | closed | 2023-12-01T14:18:37Z | 2023-12-01T14:58:23Z | 2023-12-01T14:58:22Z | albertvillanova |
2,019,200,943 | enable viewer for MERA benchmark | The viewer for the [MERA benchmark](https://huggingface.co/datasets/ai-forever/MERA) is not working because it is a script-based dataset but this is a cool benchmark for LLMs evaluation in Russian and I believe it's okay to add it to the allow list for now (along with encouraging the authors to push it as parquet in th... | enable viewer for MERA benchmark: The viewer for the [MERA benchmark](https://huggingface.co/datasets/ai-forever/MERA) is not working because it is a script-based dataset but this is a cool benchmark for LLMs evaluation in Russian and I believe it's okay to add it to the allow list for now (along with encouraging the a... | closed | 2023-11-30T17:44:33Z | 2023-12-01T15:27:33Z | 2023-11-30T17:49:32Z | polinaeterna |
2,019,136,886 | remove finished jobs | See https://github.com/huggingface/datasets-server/pull/2137#issue-2005097421 for context.
The next waiting job operation could be affected because of useless job records (finished). This PR proposes removing the jobs in SUCCESS, ERROR, and CANCELLED status so that the query can improve performance and profile better ... | remove finished jobs: See https://github.com/huggingface/datasets-server/pull/2137#issue-2005097421 for context.
The next waiting job operation could be affected because of useless job records (finished). This PR proposes removing the jobs in SUCCESS, ERROR, and CANCELLED status so that the query can improve performan... | closed | 2023-11-30T17:17:24Z | 2023-12-06T15:22:02Z | 2023-12-06T15:22:00Z | AndreaFrancis |
2,018,656,382 | allow script on another dataset | null | allow script on another dataset: | closed | 2023-11-30T13:20:01Z | 2023-11-30T13:20:41Z | 2023-11-30T13:20:08Z | severo |
2,018,367,781 | Allow script for poloclub/diffusiondb | null | Allow script for poloclub/diffusiondb: | closed | 2023-11-30T10:46:30Z | 2023-11-30T10:47:33Z | 2023-11-30T10:47:32Z | lhoestq |
2,018,255,160 | Sort __hf_index_id column | Sort `__hf_index_id` column.
After investigation, ~~I think~~ the root cause of the unsorted column is that `duckdb` uses some kind of parallelism when generating a table from a Parquet file (~~maybe~~ across row groups), calling the 'serial' sequence concurrently. To make the column sorted, I propose to add it afte... | Sort __hf_index_id column: Sort `__hf_index_id` column.
After investigation, ~~I think~~ the root cause of the unsorted column is that `duckdb` uses some kind of parallelism when generating a table from a Parquet file (~~maybe~~ across row groups), calling the 'serial' sequence concurrently. To make the column sorte... | closed | 2023-11-30T09:44:15Z | 2023-11-30T16:28:52Z | 2023-11-30T16:13:53Z | albertvillanova |
2,017,364,615 | test priority small datasets vs big dataset | Add a test to get some clues about the next waiting job issue.
- From comment https://github.com/huggingface/datasets-server/pull/2141#issuecomment-1829822409, I couldn't reproduce the issue regarding small datasets being blocked by big ones.
- I was able to evaluate and measure the delay due to reading concurrency f... | test priority small datasets vs big dataset: Add a test to get some clues about the next waiting job issue.
- From comment https://github.com/huggingface/datasets-server/pull/2141#issuecomment-1829822409, I couldn't reproduce the issue regarding small datasets being blocked by big ones.
- I was able to evaluate and m... | closed | 2023-11-29T20:27:54Z | 2024-01-26T11:55:57Z | 2023-12-14T15:28:58Z | AndreaFrancis |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.