Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

develop #1

Merged
merged 249 commits into from
Dec 29, 2024
Merged

develop #1

merged 249 commits into from
Dec 29, 2024

Conversation

meslubi2021
Copy link
Collaborator

Description of the change

Motivation

Type of change

Choose one: (Bug fix | Feature request | Documentation | Other)

Checklist

  • I have performed a self-review of my code.
  • I have added detailed comments to my code where applicable.
  • I have verified that my change does not break existing code.
  • My PR is based on the latest changes of the main branch (if unsure, please run git pull --rebase upstream main).
  • I am familiar with the Google Style Guide for the language I have coded in.
  • I have read through the Contributing Guide and signed the Contributor License Agreement.

MarkDaoust and others added 30 commits October 10, 2023 11:58
@MarkDaoust
Add dataclass prettyprinting. (#73)
e19b8b0 
* Add dataclass prettyprinting.
* use reprlib.recursive_repr
* format
* Improve contractions.
@MarkDaoust
Add metadata handling. (#74)
46db06f 
* Add metadata handling.

* Add and fix tests.

* Resolve comments
@MarkDaoust
Add count_text_tokens, and expose operations. (#76)
125c436 
* Add count_text_tokens, and expose operations.

* Format and fix pytype errors.

* use get_base_model_name in create_tuned_model

* Resolve comments
@sbagri
Update gspread_client.py (#82)
923d372 
Inside google, we  are removing support for `from gspread import gspread`
@MarkDaoust
Update dataset construction code to handle more input types. (#79)
2eb5e82 
* Allow dicts and pd.DataFrames as tuning datasets

* fix type hints.

* Support csv and json files, csv URLs.

* docs

* add TODO

* Allow json-urls and streaming decoding of CSVs.
@MarkDaoust
Restore ModelNameOptions for backwards compatibility. (#87)
2e9f23e
@markmcd
Support specific tabs when loading Google Sheets (#88)
14915d9 
The CSV URL trick seems to support specifying tabs, so this change
plumbs the tab reference (`gid`) through.
@MarkDaoust
Handle max batch size for embeddings. (#83)
0d8d339
@MarkDaoust
Simplify list-models iteration - the GAPIC lib handles this. (#86)
db0fe2d 
* Simplify list-models iteration - the GAPIC lib handles this.

* resolve comments
@MarkDaoust
Support python 3.12 (#89)
c4199bb
@MarkDaoust
Revert "Support python 3.12 (#89)" (#91)
f619444 
This reverts commit c4199bb.
@MarkDaoust
Release 0.2.2 (#90)
e5f4463 
* Revert "Support python 3.12 (#89)"

This reverts commit c4199bb.

* Create v0.2.2 release

* Update RELEASE.md - No 3.12 support.
@markmcd
Fix source URLs for glm.* members. (#94)
152d2cf 
GLAPI members defined in a versioned package (e.g. `v1beta3) were
generating an incorrect URL for the "view on github" link.

To test the regex I ran it on a suspended instance of the script:

```
diff -u <(egrep -Iir 'generativelanguage_\w+\.types' docs/api/google/ai/generativelanguage/ --include=*.md) \
        <(egrep -Iir 'generativelanguage_\w+.types' docs/api/google/ai/generativelanguage/ --include=*.md)
```

Output [here](http://go/paste/6300090913259520), but it only matched
`<a href />`s.
@markmcd
Allow refresh() to take args (#95)
e108d23 
`refresh()` can be called with args, so accept arbitrary args for the fake implementation.

This was an issue internally, where refresh appears to be called with args.
@keertk
Clean up labels + automation (#100)
5b0b406 
* Update stale.yml

* Update labeler.yml

* Update bug_report.yml

* Update feature_request.yml
@MarkDaoust @shilpakancharla @markmcd
Update to include the generative service. (#105)
0988543 
* Start code for generative service.
This is working with v1beta.

Change-Id: I96d2d4f773db8e4be621b40697016d1acbe24903

* Embedding content functionality for v1beta.

Change-Id: I1bf2379b33a607b0bb2cf77066073166a6fe9f95

* Add py3.9 support, Fix roles, UsageMetadata, and more.

Change-Id: I5fa91f191eb5eebb0fd160325b102bfb48ef8e27

* Add async suppoort. Fix some types. Add count_tokens.

Change-Id: Ic03d9caa9996843c0ac1438f52c8b10a08fd6563

* Docstrings

Change-Id: Ia2adca04a2f65bfd4ea40eec0cc18c4c98703b9f

* Add missing async type to export list.

Change-Id: I22e0aba1a59997fef6263105413b5626f1cc5a51

* Add async tests, kwargs, format.

Change-Id: Ia2c3260efe58f48183458e997ba560cc07f4b442

* docs

Change-Id: I45d2f405fb058f25fc99c30c79221f6461ebe945

* debug tests

* Add GenerationConfig at the top level

* test

* replace __init__.py

* remove -e

* drop notebook tests for now

* Update version.

* format + pytype

* Fix pytype.

* Fix tests + pytype

---------

Co-authored-by: Shilpa Kancharla <[email protected]>
Co-authored-by: Mark McDonald <[email protected]>
@MarkDaoust
Update docstrings, fix jpeg support. (#108)
4f05500 
* Update docstrings, add png support.

* black

* Improve typechecking in docs
@MarkDaoust @markmcd
Update README.md (#107)
0cd555b 
* Update README.md

* Update README.md

* Updated Gemini sample code in README

* Update docs/build_docs.py

---------

Co-authored-by: Mark McDonald <[email protected]>
@MarkDaoust
Docstring (#110)
07fb91c 
* Update docstring

* update docstrings

* 0.3.1
@markmcd
Fix MM snippet in README (#111)
c6688b7
@Andy963
fix TypeError in count_tokens & count_tokens_async (#123)
2bdec6a 
async def count_tokens(
        self,
        request: Optional[Union[generative_service.CountTokensRequest, dict]] = None,
        *,
        model: Optional[str] = None,
        contents: Optional[MutableSequence[content.Content]] = None,
        retry: OptionalRetry = gapic_v1.method.DEFAULT,
        timeout: Union[float, object] = gapic_v1.method.DEFAULT,
        metadata: Sequence[Tuple[str, str]] = (),
    if you don't use keyword args : model=model, contents=contents,  but position args, this will lead to : TypeError: count_tokens() takes from 1 to 2 positional arguments but 3 were given
@MarkDaoust
Fixup (#128)
b72d107 
* Add smkoe tests for count_tokens.

* fix docstring
@MarkDaoust
Add basic tools support. (#120)
b97aa55
@luo-anthony
Fix Javscript -> Python typo in README.md (#132)
718d716
@thesanju
Corrected some typos (#131)
f437f24
@mpursley
fix: update links to https://github.com/google/generative-ai-docs in …
b12aef7 
…the readme (#129)

* Update links to https://github.com/google/generative-ai-docs

* Update links to https://github.com/google/generative-ai-docs
@markmcd
Test PRs when project metadata changes (#121)
106daab
@Andy963 @MarkDaoust
aiter, anext not exist error (#127)
8f05227 
* aiter, anext not exist error

* Un-skip test, format with black.

* Fix aiter & anext implementations.

---------

Co-authored-by: Mark Daoust <[email protected]>
@Hamza-nabil
docs: Correct GenerateContentResponse type documentation, remove asyn…
15bc434 
…c keyword (#140)
@Andy963 @MarkDaoust
Fix: List Index out out range with empty parts [] (#138)
0a9596f 
* Fix: Index out out range with empty parts []

while question was blocked by api, the parts is an empty list [], 
so the exp: 
if len(parts) > 1 or "text" not in parts[0]:
will raise list index out of range, but not ValueError

* simplify parts length check.

* Update error message.

---------

Co-authored-by: Mark Daoust <[email protected]>
MarkDaoust and others added 24 commits September 10, 2024 09:53
@MarkDaoust
update generativelanguage version (#547)
f03ef2d 
Change-Id: Iff0d88eef997ba3ea3ede835b148831aee1213fc
@MarkDaoust
Expand .text error descriptions. (#527)
1ccbc5a 
* Expand error descriptions for #170

Change-Id: I8599dafc9e5084a43f2ce482644d0e9e16b61061

* Fix test failure cause by upgrade to protobuf>=5.0.0

Change-Id: I16a3c1964284d16efb48073662901454d4e4a6a1

* Format

Change-Id: Iee130a7e58f2cfbc1001808ac892f119338626eb
@MarkDaoust
Fix default_metadata for files API. (#548)
90f8094 
* Fix default_metadata for files API.

Change-Id: Ibba74ccda137fe4ac62525868e036a0a0d612348

* Pass the metadata through to the discovert URL

Change-Id: If8f938a1026d46aafc617d1c0e95360c463f0ba0

* fix tests, getmembers doesn't return classmethod/staticmetnod objects.

Change-Id: I9449c6f3299f820fea4128cb598f39cf359cd9ef

* format

Change-Id: I4133647fbea3b6b995ed6edf56bd2a243c2131af

* Add notes describing interaction of the discovery URL and API endpoint.

Change-Id: Ibb8740fd548cc8a8a75b32508d6d05aa6e89f633

* note abput default-metadata

Change-Id: I9647c47718d7b6981692b484b9e9b2ed0f9a42cc

* remove unused import

Change-Id: Ie49216fef5eb13e202fd276fff62e7e20f03b062
@MarkDaoust
update version (#554)
c039446 
Change-Id: Idb2f2006cdf4b37d7fc8c73112dba3a125edf175
@markmcd
Ensure tests are runnable when standalone (#552)
4f42118 
When we run tests internally that don't invoke the test class, they pass
but don't run the tests.

Our GitHub environment tests them fine but this change ensures they run
in both environments.
@MarkDaoust
Add rev14 parameters and fixes. (#561)
36e001a 
* Add rev14 parameters

Change-Id: I16f2b1f5820a6cf867b9abb04ffd5c6e6d2d947b

* Fix flakey repr test

Change-Id: I89bcf1494cf72c6ee28f2b52d0345cbb40859862

* format

Change-Id: I81cff23e9ce20cc20b4a0632d557c71f536fd485

* Use client preview

Change-Id: I2d8a4ee2e9e4b6e00a16a9dac1136a2fa18d7a28

* Fix tests

Change-Id: If8fbbba1966aa42601adec877e60d851d4f03b72

* Fix tuned model tests

Change-Id: I5ace9222954be7d903ebbdabab9efc663fa79174

* Fix tests

Change-Id: Ifa610965c5d6c38123080a7e16416ac325418285

* format

Change-Id: I15fd5701dd5c4200461a32c968fa19e375403a7e

* pytype

Change-Id: I08f74d08c4e93bbfdf353370b5dd57d8bf86a637

* pytype

Change-Id: If81b86c176008cd9a99e3b879fbd3af086ec2235

* 3.9 tests

Change-Id: I13e66016327aae0b0f3274e941bc615f379e5669
@MarkDaoust
Update to use google.generativelanguage 0.6.10. (#568)
8f7f5cb 
* use the new generativelanguage==6.10

Change-Id: Ib65e073ad61c810e5b4d3c01bce71fc45af188d3

* Update version

Change-Id: I32c344c366cfa753230435d6b4576f3309fc968f

* 0.6.10

Change-Id: I265b5a3ab6ce598ce4c6383c5b4dd065842276f3
@PicardParis @MarkDaoust
fix: preserve quality and optimize transfer of prompt images (#570)
6c8dad1 
* fix: preserve quality and optimize transfer of prompt images

* Move numpy-images to their own test case.

Change-Id: Ie6b02c7647487c1df9d4e70e9b8eed70dc8b8fe3

* Format with black

Change-Id: I04550a89eed9bb21c0a8f6f9b6ab76b8b0f41270

---------

Co-authored-by: Mark Daoust <[email protected]>
@shilpakancharla @MarkDaoust
Search grounding (#558)
d5103eb 
* Updated tests and current progress on adding search grounding.

* Update google/generativeai/types/content_types.py

Co-authored-by: Mark Daoust <[email protected]>

* Update tests/test_content.py

Co-authored-by: Mark Daoust <[email protected]>

* Update search grounding

* update content_types

* Update and add aditional test cases

* update test case on empty_dictionary_with_dynamic_retrieval_config

* Update test cases and _make_search_grounding

* fix tests

Change-Id: Ib9e19d78861da180f713e09ec93d366d5d7b5762

* Remove print statement

* Fix tuned model tests

Change-Id: I5ace9222954be7d903ebbdabab9efc663fa79174

* Fix tests

Change-Id: Ifa610965c5d6c38123080a7e16416ac325418285

* format

Change-Id: Iab48a9400d53f3cbdc5ca49c73df4f6a186a867b

* fix typing

Change-Id: If892b20ca29d1afb82c48ae1a49bef58e0421bab

* Format

Change-Id: I51a51150879adb3d4b6b00323e0d8eaf4c0b2515

---------

Co-authored-by: Mark Daoust <[email protected]>
@MarkDaoust
Remove finish_message (#571)
be00c19 
Change-Id: I55f8339aea7e71e0b073c1a2ef9b3748500e09ed
@Hamza-nabil @MarkDaoust
Stream file upload (#556)
b50341e 
* Add IOBase support to FileServiceClient.create_file

* add support to upload file-like object

* Give clear errors for 'unknown' mime-types

Change-Id: Iea071c396c4cfe2b2c8eacae74dd8fb0acbc128f

* Remove duplicate check for mime_type

* Add a test uploading from a file-like IO object

Change-Id: I572f02ed98b9ca45299b76e7a01695fdcf917e1e

* fix type check

Change-Id: I220c05eee73ae76ced25254d67332f70a4069f7e

---------

Co-authored-by: Mark Daoust <[email protected]>
@MarkDaoust
Update default model name (#576)
4354af6 
Change-Id: I2d8ef6d251d0128695e45df119c31d6b63e26ba7
@shilpakancharla @MarkDaoust
Removing OAuth from tuning for REST (#579)
8f77cc6 
* Removing OAuth from tuning for REST

* remove key

* Fix line continuation

Change-Id: I639aa453699766bf31f7722a59f4157a53cf7de4

---------

Co-authored-by: Mark Daoust <[email protected]>
@MarkDaoust
update version (#588)
7546026 
Change-Id: I998fa990bc7dfb220e78cd7f9692c8530993d7c4
@Goldziher @MarkDaoust
feat: add py.typed (#509)
c8eadc4 
* Update setup.py

* Create py.typed

* Update setup.py to trigger CI.

---------

Co-authored-by: Mark Daoust <[email protected]>
@MarkDaoust
Remove references to Pro 1.0 (#600)
e9b0cde 
* remove references to 1.0-pro

Change-Id: I405c87d495c73550cfbd00a13249cb1e30ab0989

* remove references to gemini-pro

Change-Id: Ied2f0b7112dd5d61390da3e84457a2fb3f770665

* Update models.py

* format

Change-Id: Ib3a0c90bfc6ec7f8f793917b3140769e2635a8e9
@MarkDaoust
Fix GenerationConfig to match the API. (#609)
3d91916 
* Fix GenerationConfig, this doesn't match the API

Change-Id: I4e0c3b45125023e056c3d7e1549eafb47e42815a

* format

Change-Id: I161dd7b9b371e005b4385faeaa84f6d141e09b43
@Sumeet213
fix: thread-safe discovery API setup (Issues#327) (#583)
db8221a
@westonruter
Use double quotes for JSON object keys (#612)
c16df61
@MarkDaoust
Add whitespace to generated tables to fix markdown in tables. (#602)
0e5c5f2 
* fixes 563

Change-Id: Ifb9743822c744845aac4deea9963a0c6e6389362

* Refresh docs for v0.8.3

Change-Id: I7536e809b9a1b8f26fc970a42c83000b0d607793

* remove align-left

Change-Id: I0462300f198e3de020cbfc032e575e094b88d360

* format

Change-Id: If83605c320d2b7255618737c2cd985a153584343

* refresh

Change-Id: I0776467ac6471b2f8f06f089a54e447d149b7d78

* remove devsite headers

Change-Id: I2d10299b5cffc2e353b71dacfabb3a7e9f22a9fd

* remove devsite headers

Change-Id: I1b9ca5bf2d46dbd39b8cb226d153107851e2e3f0
@vishal-dharm
Fix 'argument list too long' error and add couple vision examples (#634)
a04fcd1
@MarkDaoust
Move imports into samples (#638)
4c00257 
* Move imports into samples

Change-Id: I86550a09c6555a36679dad517ed4239f7ded90ff

* format

Change-Id: I0c72841eb2ebad0075d0024624a42fe36fef883f
@Annhiluc @vishal-dharm
fix: Pass along model version in GenerateContentResponse (#621)
b8772ed 
* fix: Pass along model_version in GenerateContentResponse.

* Revert autogenerated doc files from 94eb16e.

* Fix 'argument list too long' error and add couple vision examples (#634)

* Update google-ai-generativelanguage version in requirements.

* Format updated generation_types and test using black.

---------

Co-authored-by: Vishal Dharmadhikari <[email protected]>
@MarkDaoust
Copy AIS templates from cookbook. (#635)
5a3ba73 
https: //github.com/google-gemini/cookbook/tree/e2623e8587f3d4ae13df65f415a3337c7f7ac97e/templates
Change-Id: Ic50b279a769d2fe2263b1dee41541d7c950c0cfa
@meslubi2021 meslubi2021 self-assigned this Dec 29, 2024
@socket-security Socket Security
Copy link

New dependencies detected. Learn more about Socket for GitHub ↗︎

Package New capabilities Transitives Size Publisher
pypi/[email protected] environment, eval, filesystem, shell 0 14.9 MB aclark, hugovk, radarhere, ...1 more

View full report↗︎

@socket-security Socket Security
Copy link

Report is too large to display inline.
View full report↗︎

Next steps

Mark a package as acceptable risk

To ignore an alert, reply with a comment starting with @SocketSecurity ignore followed by a space separated list of ecosystem/package-name@version specifiers. e.g. @SocketSecurity ignore npm/[email protected] or ignore all packages with @SocketSecurity ignore-all

@meslubi2021 meslubi2021 merged commit ee62753 into Unity-google-new:main Dec 29, 2024
12 of 13 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees

@meslubi2021 meslubi2021

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.