Releases: argilla-io/argilla-server
v1.28.0
1.28.0
Added
- Added support to specify a list of score values for suggestions
score
attribute. (#98) - Added
GET /api/v1/settings
new endpoint exposing Argilla and Hugging Face settings when available. (#127) - Added
ARGILLA_SHOW_HUGGINGFACE_SPACE_PERSISTENT_STORAGE_WARNING
new environment variable to disable warning message when Hugging Face Spaces persistent storage is disabled. (#124) - Added
options_order
new settings attribute to support specify an order for options in multi label selection questions. (#133) - Added
POST /api/v1/datasets/:dataset_id/records/bulk
endpoint. (#106) - Added
PUT /api/v1/datasets/:dataset_id/records/bulk
endpoint. (#106)
Deprecated
- Deprecated
POST /api/v1/datasets/:dataset_id/records
in favour ofPOST /api/v1/datasets/:dataset_id/records/bulk
. (#130) - Deprecated
PATCH /api/v1/dataset/:dataset_id/records
in favour ofPUT /api/v1/datasets/:dataset_id/records/bulk
. (#130)
Removed
- Removed support for specifying
score
attributes for individual value items when creating suggestions associated with span questions. (#101)
Full Changelog: v1.27.0...v1.28.0
v1.27.0
1.27.0
Added
- Added
allow_overlapping
field for creation and update of span question settings. (#89) - Added
ARGILLA_LABEL_SELECTION_OPTIONS_MAX_ITEMS
environment variable to set the number of maximum items to be used by label and multi label questions. By default this value is set to500
. (#85) - Added
ARGILLA_SPAN_OPTIONS_MAX_ITEMS
environment variable to set the number of maximum items to be used by span questions. By default this value is set to500
. (#85) - Added
GET /api/v1/datasets/:dataset_id/progress
endpoint to return progress metrics related with one specific dataset. (#90)
Changed
- Changed ElasticSearch JVM heap size from
512m
to1g
for quickstart Dockerfile. (#109)
Full Changelog: v1.26.1...v1.27.0
v1.26.1
1.26.1
Note
This patch version only includes changes in the argilla frontend to support RTL languages.
No changes where made on argilla-server
repository for this patch version.
Added
- Added latests changes from argilla to support for automatic detection of RTL languages. (#4686)
Full Changelog: v1.26.0...v1.26.1
v1.26.0
1.26.0
Added
- Add support for new
span
questions. (#54) - Add
inserted_at
andupdated_at
missing fields to API v1Suggestion
schema. (#52)
Fixed
- Fixed error when returning responses from deleted users (which contains user_id=None). (#57)
New Contributors
- @davidberenstein1957 made their first contribution in #41
Full Changelog: v1.25.0...v1.26.0
v1.25.0
1.25.0
Important
This version includes changes related to the search index. So, a reindex is needed.
Visit the docs for more info.
Added
- Added
ARGILLA_ES_MAPPING_TOTAL_FIELDS_LIMIT
to work with large dataset annotation flows. (#31) - Added support for update options for label and multi-label selection questions. (#34)
- Added
REINDEX_DATASETS
quickstart environment variable to reindex datasets and records into the search engine. (#39) - Added pseudo-random shuffling of records based on the user's
id
. (#23)
Changed
- Changed indexing responses for search to use user
id
instead ofusername
. Reindex needed (#26) - Changed search index mappings definition to optimize the number of fields. Reindex needed (#31)
Removed
- Removed
ARGILLA_LOCAL_AUTH_TOKEN_EXPIRATION_IN_MINUTES
environment variable. (#38) - Removed
ARGILLA_LOCAL_AUTH_ALGORITHM
environment variable. (#38) - Removed
ARGILLA_LOCAL_AUTH_SECRET_KEY
environment variable. (#38)
Fixed
- Max size parameter for getting the metadata property metrics is currently set as 2^14(=12) instead of 2 ** 14 (#30) (v1.24-fix)
- Fixed error when combining similarity search with text search. (#32)
New Contributors
- @bharath97-git made their first contribution in #30
- @gabrielmbmb made their first contribution in #23
Full Changelog: v1.24.0...v1.25.0
v1.25.0.rc1
What's Changed
- chore: review the CHANGELOG.md file by @frascuchon in #42
Full Changelog: v1.25.0.rc0...v1.25.0.rc1
v1.25.0.rc0
What's Changed
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #16
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #25
- fix: test suite is now fully passing using SQLite or PostgreSQL by @jfcalvo in #27
- feat: using user
id
for index responses in search indexes by @frascuchon in #26 - fix: max size parameter is configured to 2 ** 14 instead of 2 ^ 14 by @bharath97-git in #30
- fix: combine text with similarity search by @frascuchon in #32
- refactor: Separate versioned API endpoints by @frascuchon in #28
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #29
- feat: add dotenv files for PDM custom commands by @jfcalvo in #33
- feat: add different deterministic sorting of records per user by @gabrielmbmb in #23
- feat: Allow update question labels by @frascuchon in #34
- feat: Allow increase total number of fields for large dataset annotation flows. by @frascuchon in #31
- remove backward compatibility with argilla local auth env variables by @frascuchon in #38
- feat: Allow reindex datasets on quickstart server startup by @frascuchon in #39
New Contributors
- @bharath97-git made their first contribution in #30
- @gabrielmbmb made their first contribution in #23
Full Changelog: v1.24.0...v1.25.0.rc0
v1.24.0
1.24.0
Note
This version is the first release of the Argilla Server. Before this release, the Argilla Server was part of the Argilla SDK.
Now, the Argilla Server is a separate package that can be installed and used independently of the Argilla SDK.
Fixed
- Fixed problems using
ARGILLA_BASE_URL
environment variable. (#14)
Full Changelog: https://github.com/argilla-io/argilla-server/commits/v1.24.0