search

GoogleCloudPlatform / genai-databases-retrieval-app

Apache License 2.0
212 stars 72 forks source link

deps: Update langchain #492

Open renovate-bot opened 1 month ago

renovate-bot commented 1 month ago

This PR contains the following updates:

Package Change Age Adoption Passing Confidence
google-cloud-aiplatform ==1.60.0 -> ==1.66.0 age adoption passing confidence
google-cloud-aiplatform ==1.62.0 -> ==1.66.0 age adoption passing confidence
langchain (changelog) ==0.2.11 -> ==0.3.0 age adoption passing confidence
langchain-community (changelog) ==0.2.9 -> ==0.3.0 age adoption passing confidence
langchain-core (changelog) ==0.2.24 -> ==0.3.1 age adoption passing confidence
langchain-google-vertexai ==1.0.7 -> ==1.0.10 age adoption passing confidence
langchain-text-splitters (changelog) ==0.2.2 -> ==0.3.0 age adoption passing confidence

Release Notes

googleapis/python-aiplatform (google-cloud-aiplatform) ### [`v1.66.0`](https://redirect.github.com/googleapis/python-aiplatform/blob/HEAD/CHANGELOG.md#1660-2024-09-11) [Compare Source](https://redirect.github.com/googleapis/python-aiplatform/compare/v1.65.0...v1.66.0) ##### Features - Add max_wait_duration option to custom jobs. ([ee65917](https://redirect.github.com/googleapis/python-aiplatform/commit/ee65917fbc1edc58a7e57c3601f9329ffa1304b0)) - Add Ray 2.33 support to SDK Client Builder, remove deprecated protocol_version from ray client context. ([708a67b](https://redirect.github.com/googleapis/python-aiplatform/commit/708a67b9b781d41564058904d60af7db9beafbe0)) - Add support for version 2.33 for RoV Bigquery read/write, remove dead code from version 2.4 ([a4b6c60](https://redirect.github.com/googleapis/python-aiplatform/commit/a4b6c60051efe61ad590ede8682bd36b79cafc94)) - Update Ray system tests to be compatible with new RoV 2.33 changes ([8c7bf27](https://redirect.github.com/googleapis/python-aiplatform/commit/8c7bf2724037f1ad1081d29e9534f8d26b18d95f)) - Update setup.py for including Ray v2.33, restrict RoV prediction to 2.9.3 for now ([71c6f3c](https://redirect.github.com/googleapis/python-aiplatform/commit/71c6f3c6b48a001bfdbe5f5b47d787d4797ae9bf)) ##### Bug Fixes - Identify nested Vertex Tensorboard profile runs for uploading when standard event data is not present ([0a08027](https://redirect.github.com/googleapis/python-aiplatform/commit/0a08027df7a7442af63522039445500ce1fdb7f1)) - Move region tag to include imports ([6d1f7fd](https://redirect.github.com/googleapis/python-aiplatform/commit/6d1f7fdaadade0f9f6a77c136490fac58d054ca8)) - Move region tag to include imports ([#​4357](https://redirect.github.com/googleapis/python-aiplatform/issues/4357)) ([6d1f7fd](https://redirect.github.com/googleapis/python-aiplatform/commit/6d1f7fdaadade0f9f6a77c136490fac58d054ca8)) - Rollback change to tensorboard uploader causing increased latency ([879dbcd](https://redirect.github.com/googleapis/python-aiplatform/commit/879dbcd1ed49bee573df65e764914b708382404c)) ##### Documentation - Update SDK Job Submission with Ray v2.33 updated comments ([7fda11f](https://redirect.github.com/googleapis/python-aiplatform/commit/7fda11fd25b5a78fcca69e48f9c734f2a8f86eb0)) ### [`v1.65.0`](https://redirect.github.com/googleapis/python-aiplatform/blob/HEAD/CHANGELOG.md#1650-2024-09-04) [Compare Source](https://redirect.github.com/googleapis/python-aiplatform/compare/v1.64.0...v1.65.0) ##### ⚠ BREAKING CHANGES - Tokenization - Deprecated `ComputeTokenResult.token_info_list` in favor of `ComputeTokenResult.tokens_info` ##### Features - Add support for system instruction and tools in tokenization. ([72fcc06](https://redirect.github.com/googleapis/python-aiplatform/commit/72fcc063ed4a086da0ad37ec2ac58860d4e79051)) - Add vector search alpha to rag retrieval for hybrid search ranking ([6624ebe](https://redirect.github.com/googleapis/python-aiplatform/commit/6624ebe22726942dd70781122352f47268ee2dee)) - Adding Weaviate Vector DB option for RAG corpuses to SDK ([9b28202](https://redirect.github.com/googleapis/python-aiplatform/commit/9b28202a1bb17f54e042301d6cdac3b6aa826797)) - GenAI - Added `system_instruction` and `tools` support to `GenerativeModel.count_tokens` ([50fca69](https://redirect.github.com/googleapis/python-aiplatform/commit/50fca693b2f3b1a0b61867dc136be5a468fb2b2f)) - GenAI - Added Llama3 support in GenAI batch prediction ([6166152](https://redirect.github.com/googleapis/python-aiplatform/commit/6166152844dc0078f7a5a02355ef3555cc428cfa)) - PrivateEndpoint.stream_raw_predict ([197f333](https://redirect.github.com/googleapis/python-aiplatform/commit/197f333be5a075d41f98b762cd933cd2e89cecae)) - Support reserved_ip_ranges for VPC network in Ray on Vertex cluster ([36a56b9](https://redirect.github.com/googleapis/python-aiplatform/commit/36a56b99f9e53d19d80c2bff3bf55c208988c518)) - Tokenization - Deprecated `ComputeTokenResult.token_info_list` in favor of `ComputeTokenResult.tokens_info` ([efbcb54](https://redirect.github.com/googleapis/python-aiplatform/commit/efbcb54e0d5df4d65a79e60afdbc5c328538aef6)) ##### Bug Fixes - Tensorboard - Fixed bug in tensorboard uploader where profile logs are not uploaded from nested run directories. ([37627de](https://redirect.github.com/googleapis/python-aiplatform/commit/37627de4ec12b8c51abf41524b9e7e3adf2dab54)) - Tokenizers - Fixed `Tokenizer.compute_tokens` ([c29fa5d](https://redirect.github.com/googleapis/python-aiplatform/commit/c29fa5d98fc1c1550c039e84ff2f5725818c2231)) ### [`v1.64.0`](https://redirect.github.com/googleapis/python-aiplatform/blob/HEAD/CHANGELOG.md#1640-2024-08-27) [Compare Source](https://redirect.github.com/googleapis/python-aiplatform/compare/v1.63.0...v1.64.0) ##### Features - Endpoint - Add support for Prediction dedicated endpoint. predict/rawPredict/streamRawPredict can use dedicated DNS to access the dedicated endpoint. ([3d68777](https://redirect.github.com/googleapis/python-aiplatform/commit/3d687777d39b00280c22d2a14ddde3ba644febf7)) - GenAI - Added the model Distillation feature (private preview) ([a0d4ff2](https://redirect.github.com/googleapis/python-aiplatform/commit/a0d4ff20ceb1c48806d1711fdb2691dc34f9f1db)) - Grounding - Allow initialization of `grounding.VertexAISearch` with full resource name or data store ID, project ID, and location. ([f334321](https://redirect.github.com/googleapis/python-aiplatform/commit/f334321694bb3be1a421ee19a33fc973e5455da9)) - Evaluation - Make Rouge class available in base level init ([aed82a1](https://redirect.github.com/googleapis/python-aiplatform/commit/aed82a1bd5d8942ded4dd325a4eb1a5f73bc50c7)) - Feature Store - Read for online store w/private service connect ([7af80c6](https://redirect.github.com/googleapis/python-aiplatform/commit/7af80c624e026cfe8dda2d1644467a517f49b78f)) - Ray - Support autoscaling in Ray on Vertex ([961da42](https://redirect.github.com/googleapis/python-aiplatform/commit/961da429683db113a822300342484aaf1128cfc0)) ##### Bug Fixes - Fix error in tensorboard uploader thrown when time_series_id is None ([d59a052](https://redirect.github.com/googleapis/python-aiplatform/commit/d59a0522ddc2131ab39b052e742a6472f84e0a5a)) - Evaluation - Fix typo in prompt templates: ([c8fa7a8](https://redirect.github.com/googleapis/python-aiplatform/commit/c8fa7a8cf53165354fa89e38fffc0ef4a821e211)) ##### Documentation - **samples:** Adding code sample for vector search create streaming index ([71464e7](https://redirect.github.com/googleapis/python-aiplatform/commit/71464e7d2a57fa6770d2fcb7c5c0e669055c4cdb)) ### [`v1.63.0`](https://redirect.github.com/googleapis/python-aiplatform/blob/HEAD/CHANGELOG.md#1630-2024-08-20) [Compare Source](https://redirect.github.com/googleapis/python-aiplatform/compare/v1.62.0...v1.63.0) ##### Features - A new field `satisfies_pzs` is added to message `.google.cloud.aiplatform.v1.BatchPredictionJob` ([#​4192](https://redirect.github.com/googleapis/python-aiplatform/issues/4192)) ([6919037](https://redirect.github.com/googleapis/python-aiplatform/commit/6919037e9513e922e9ffe197e68a99fb343c4fff)) - Add advanced PDF parsing option for RAG file import ([6e1dc06](https://redirect.github.com/googleapis/python-aiplatform/commit/6e1dc0658ffd875f4a3bbcab62976e15e997102e)) - Add multithreading for custom metric computation. ([2c93fc1](https://redirect.github.com/googleapis/python-aiplatform/commit/2c93fc17b4a76623209b7699a73d4e6c9c27cc81)) - Add progress bar for generating inference. ([b78714f](https://redirect.github.com/googleapis/python-aiplatform/commit/b78714f3cf8fc22f7caa193a7398efe3626c2c5f)) - Add progress bar to custom metrics. ([3974aec](https://redirect.github.com/googleapis/python-aiplatform/commit/3974aec92595870b6f33ecd016763f59d6630898)) - Add Prompt class support for configs and Prompt.generate_content wrapper ([7f1e031](https://redirect.github.com/googleapis/python-aiplatform/commit/7f1e0313842546b7e911ee3ef06d7193deb64a91)) - GenAI - Added seed parameter to the GenerationConfig class ([9f1e073](https://redirect.github.com/googleapis/python-aiplatform/commit/9f1e0739d837b3110b40b8806514ca3e49e2b1da)) - GenAI - Added the `Candidate.avg_logprobs` property ([de80695](https://redirect.github.com/googleapis/python-aiplatform/commit/de80695ad2359361a698cffbde2336417297ef35)) - GenAI - Released the `Prompt` feature to Public Preview ([64eeab8](https://redirect.github.com/googleapis/python-aiplatform/commit/64eeab8b3404e87c0cc19fb6862ff51ec8b95954)) - GenAI Evaluation: Add generic model-based `PointwiseMetric` and `PairwiseMetric` classes that allow customizing metric prompt templates. Add `PointwiseMetricPromptTemplate`, `PairwiseMetricPromptTemplate` classes to help formulate and customize metric prompt templates. Add `metric_column_mapping` parameter to `EvalTask` for metric prompt template input variable name mapping. ([fd38b49](https://redirect.github.com/googleapis/python-aiplatform/commit/fd38b49231bd1b35af57056b9e69a5427ddb114c)) - GenAI Evaluation: Open source model-based metric prompt template examples for Gemini 1.5 Pro autorater. Add `MetricPromptTemplateExamples` class to help retrieve model-based metric prompt templates. ([fd38b49](https://redirect.github.com/googleapis/python-aiplatform/commit/fd38b49231bd1b35af57056b9e69a5427ddb114c)) - GenAI Evaluation: Release GenAI Evaluation SDK GA features to `vertexai.preview` module. ([fd38b49](https://redirect.github.com/googleapis/python-aiplatform/commit/fd38b49231bd1b35af57056b9e69a5427ddb114c)) - Publish GenAI Evaluation SDK GA features to `vertexai.evaluation` module. Switch GenAI Evaluation Service client to v1 version. ([45e4251](https://redirect.github.com/googleapis/python-aiplatform/commit/45e42516fbc47db1c44a7669f2730a1590a992c2)) ##### Bug Fixes - Add support of display_name to create_cached_content in python SDK ([ecc2d54](https://redirect.github.com/googleapis/python-aiplatform/commit/ecc2d54a84c03f7d06e987fba5f5c67fba109ce0)) - Always upload the pickled object and dependencies tarball when creating ReasoningEngine ([34ef5a3](https://redirect.github.com/googleapis/python-aiplatform/commit/34ef5a35bd91aea53c89650c20962dae29b3a535)) - Remove grouding attribution ([f6ece65](https://redirect.github.com/googleapis/python-aiplatform/commit/f6ece65d8f2933ddcb4ec1a08784f8d2c365f2de)) ##### Documentation - Update Prompt docstring for batch prompting ([e96b6e6](https://redirect.github.com/googleapis/python-aiplatform/commit/e96b6e6f0574b96fc4b61f99c671ef6646fc4956)) ### [`v1.62.0`](https://redirect.github.com/googleapis/python-aiplatform/blob/HEAD/CHANGELOG.md#1620-2024-08-13) [Compare Source](https://redirect.github.com/googleapis/python-aiplatform/compare/v1.61.0...v1.62.0) ##### Features - Add metadata to evaluation result. ([375095e](https://redirect.github.com/googleapis/python-aiplatform/commit/375095e72cc4f43611710372a1e36753a891a710)) - Add Prompt class for multimodal prompt templating ([1bdc235](https://redirect.github.com/googleapis/python-aiplatform/commit/1bdc235ea64f8d63ce9d60d88cb873ee341d3ff9)) - Add support for query method in Vertex AI Extension SDK ([0008735](https://redirect.github.com/googleapis/python-aiplatform/commit/0008735968606a716add88072cff76f2fc552d7b)) - Add support for reservation affinity in custom training jobs. ([802609b](https://redirect.github.com/googleapis/python-aiplatform/commit/802609b1f5e5d8d41a77dafb5b1a2dbf01f2bd30)) - Add support for strategy in custom training jobs. ([a076191](https://redirect.github.com/googleapis/python-aiplatform/commit/a076191b8726363e1f7c47ef8343eb86cebf9918)) - Adding spot, reservation_affinity to Vertex SDK ([3e785bd](https://redirect.github.com/googleapis/python-aiplatform/commit/3e785bd9c9d3d11197ef930f563ee96231a67d84)) - Support api keys in initializer and create_client ([7404f67](https://redirect.github.com/googleapis/python-aiplatform/commit/7404f679246e41e0009ec2d49f05d669eb357f71)) - Support creating optimized online store with private service connect ([659ba3f](https://redirect.github.com/googleapis/python-aiplatform/commit/659ba3f287f9aa78840d4b9b9ca216002d5f1e6a)) - Support disable Cloud logging in Ray on Vertex ([accaa97](https://redirect.github.com/googleapis/python-aiplatform/commit/accaa9750d98b7a37b08da3bd2058d9cdd03bd5c)) - Support PSC-Interface in Ray on Vertex ([accaa97](https://redirect.github.com/googleapis/python-aiplatform/commit/accaa9750d98b7a37b08da3bd2058d9cdd03bd5c)) ##### Bug Fixes - Added credentials, project, and location on PipelineJobSchedule init ([281c171](https://redirect.github.com/googleapis/python-aiplatform/commit/281c1710afc6cac49c02d926bee7a6c43b6ef851)) - Avoid breakage of langchain from orjson 3.10.7 ([c990f73](https://redirect.github.com/googleapis/python-aiplatform/commit/c990f73845f38e58ba2dddb372ad2f84d4a05479)) - Deprecate disable_attribution in GoogleSearchRetrieval. ([c68d559](https://redirect.github.com/googleapis/python-aiplatform/commit/c68d559b9d0fd7288b6775f57d05f474f5f7920a)) ##### Documentation - Update the docstring for compute_tokens method. ([849e8d4](https://redirect.github.com/googleapis/python-aiplatform/commit/849e8d409e4838cad0a020231b806b0c9ef587ce)) ### [`v1.61.0`](https://redirect.github.com/googleapis/python-aiplatform/blob/HEAD/CHANGELOG.md#1610-2024-08-05) [Compare Source](https://redirect.github.com/googleapis/python-aiplatform/compare/v1.60.0...v1.61.0) ##### Features - Add a warning message for scheduled deprecation of Coherence metric class ([7f238fb](https://redirect.github.com/googleapis/python-aiplatform/commit/7f238fb3cebc44893b4e6959a77743cc4d96138e)) - Add deprecation messages for all model-based metric classes ([71c0fd3](https://redirect.github.com/googleapis/python-aiplatform/commit/71c0fd397139a95b6045f898e906ce11b2e7e8ce)) - Add support for task type (CODE_RETRIEVAL_QUERY) through get_embeddings. ([f2ce1e4](https://redirect.github.com/googleapis/python-aiplatform/commit/f2ce1e4caea9f344e39fc3232f697b1a6ea4f99a)) - Add system_instruction to LangchainAgent template. ([c71c3dd](https://redirect.github.com/googleapis/python-aiplatform/commit/c71c3ddbfeaa577dfce683b3299d94e77d1c4895)) - Adding Slack and Jira data connector for RAG to SDK ([d92e7c9](https://redirect.github.com/googleapis/python-aiplatform/commit/d92e7c91d280dd417d2c2a2cf5abc36592888593)) - Allow protobuf 5.x ([ce9cd5d](https://redirect.github.com/googleapis/python-aiplatform/commit/ce9cd5def14597822c1d071e438cf63b6d4ba3ca)) - LVM - Release `ImageGenerationModel` to GA ([718c199](https://redirect.github.com/googleapis/python-aiplatform/commit/718c1997778310b6898344b2e5a34513e7a82e5f)) - Support "update" for reasoning engine. ([b73ef3e](https://redirect.github.com/googleapis/python-aiplatform/commit/b73ef3eaa2d88dbc8071e3a4f0c7da934683fc2a)) - Update Rapid Evaluation Service QPS. Add a customizable evaluation service QPS parameter. ([9ee9289](https://redirect.github.com/googleapis/python-aiplatform/commit/9ee9289fbe5face719515e453d4f81648b44e7b1)) ##### Documentation - Change init sample to use vertexai ([829e0f6](https://redirect.github.com/googleapis/python-aiplatform/commit/829e0f6fd286cf2de2ac307a836305766473faef)) - Make small fixes to file import documentation ([f7d65c3](https://redirect.github.com/googleapis/python-aiplatform/commit/f7d65c32948c54bcf3a6927639f2173b556bb310))

Configuration

📅 Schedule: Branch creation - At any time (no schedule defined), Automerge - At any time (no schedule defined).

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Never, or you tick the rebase/retry checkbox.

👻 Immortal: This PR will be recreated if closed unmerged. Get config help if that's undesired.

This PR was generated by Mend Renovate. View the repository job log.

dpebot commented 1 month ago

/gcbrun

kurtisvg commented 1 month ago

Probably need to do #489 in this PR

anubhav756 commented 1 month ago

490 updates the LLM demo's modules (including langchain and langchain-google-vertexai) with the necessary code changes. In that PR, I'll also address the lint and evaluation issues and also pin the module versions to the current latest. I can then apply these fixes to the retrieval service as well.

dpebot commented 1 month ago

/gcbrun

dpebot commented 1 month ago

/gcbrun

dpebot commented 1 month ago

/gcbrun

dpebot commented 1 month ago

/gcbrun

dpebot commented 1 month ago

/gcbrun

dpebot commented 1 month ago

/gcbrun

dpebot commented 1 month ago

/gcbrun

forking-renovate[bot] commented 1 month ago

Edited/Blocked Notification

Renovate will not automatically rebase this PR, because it does not recognize the last commit author and assumes somebody else may have edited the PR.

You can manually request rebase by checking the rebase/retry box above.

⚠️ Warning: custom changes will be lost.