vertexai

1.70.0 (2024-10-08)

Features

• Add a dynamic retrieval API (44df243)
• Add enable_secure_private_service_connect in service attachment (44df243)
• Add new PscInterfaceConfig field to pipeline_job.proto (44df243)
• Add psc_automation_configs to DeployIndex v1 (44df243)
• Add TunedModelRef and RebaseTunedModel Api for Vertex GenAiTuningService (44df243)
• Add TunedModelRef and RebaseTunedModel Api for Vertex GenAiTuningService (44df243)
• Add update_corpus method for vertex rag (09353cf)
• Automatically end Experiment runs when Tensorboard CustomJob is complete (30cf221)
• Vision Models - Add new safety filter levels to Vertex AI Image Generation API. (b1d5007)

Bug Fixes

• Annotate PipelineJob and PipelineTaskRerunConfig fields as optional (44df243)
• Update the docstring for LangchainAgent (f7fe2b0)

Documentation

• Add sample to retrieve experiment backing tensorboard resource name (138dc1a)
• GenAI - update comments in GenerationConfig property (c670eeb)
• Limit comment SupervisedTuningSpec for 1p tuning (44df243)

1.69.0 (2024-10-01)

Features

• Add rerun method to pipeline job preview client. (29dec74)

Bug Fixes

• GenAI - Fixed from_dict methods (3090812)

1.68.0 (2024-09-24)

Features

• A new field response_logprbs is added to message .google.cloud.aiplatform.v1.GenerationConfig (#4410) (470933f)
• Add support for partial failures sink in import rag files. (8070411)
• Add support for SharePoint as a ImportRagFiles source. (f89df1f)
• Adding Vertex Vector Search Vector DB option for RAG corpuses to SDK (d4193f2)
• Batch_predict method generally-available at TextEmbeddingModel. (73c0dae)
• GenAI - Add model routing config to sdk (c0626fe)
• GenAI - Add support for logprobs and response_logprobs. (7acf0f7)

Bug Fixes

• Fix rag corpus creation error (d25edce)
• Fix typo in eval_task docstring (86fc215)
• Fix typos in evaluation example metric prompt templates. (5f4d586)
• GenAI - Fixed GenerativeModel.compute_tokens for v1 API (4637b4c)
• Tensorboard - Fix error in tensorboard batch upload of nested dirs (86be328)

Documentation

• Update Gemini docs (c561aa6)
• Update Gemini docs (2b84142)

1.67.1 (2024-09-18)

Bug Fixes

• Fix rag corpus creation error (6083b93)

1.67.0 (2024-09-17)

Features

• Add support for partial failures sink in import rag files. (07e471e)
• Adding Feature Store Vector DB option for RAG corpuses to SDK (cfc3421)
• Adding Pinecone Vector DB option for RAG corpuses to SDK (f78b953)
• Adding Vertex Vector Search Vector DB option for RAG corpuses to SDK (f882657)
• Allow customizing pipeline caching options for model evaluation jobs. (73490b2)
• GenAI - Switched the GA version of the generative_models classes to use the v1 service APIs instead of v1beta1 (66d84af)

Bug Fixes

• GenAI - Fixed GenerativeModel.compute_tokens for v1 API (0de2987)
• Tensorboard - Fix error in tensorboard batch upload of nested dirs (ab157c8)

Documentation

• Manually add summary overview page. (0bc608a)

1.66.0 (2024-09-11)

Features

• Add max_wait_duration option to custom jobs. (ee65917)
• Add Ray 2.33 support to SDK Client Builder, remove deprecated protocol_version from ray client context. (708a67b)
• Add support for version 2.33 for RoV Bigquery read/write, remove dead code from version 2.4 (a4b6c60)
• Update Ray system tests to be compatible with new RoV 2.33 changes (8c7bf27)
• Update setup.py for including Ray v2.33, restrict RoV prediction to 2.9.3 for now (71c6f3c)

Bug Fixes

• Identify nested Vertex Tensorboard profile runs for uploading when standard event data is not present (0a08027)
• Move region tag to include imports (6d1f7fd)
• Move region tag to include imports (#4357) (6d1f7fd)
• Rollback change to tensorboard uploader causing increased latency (879dbcd)

Documentation

• Update SDK Job Submission with Ray v2.33 updated comments (7fda11f)

1.65.0 (2024-09-04)

⚠ 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)
• Add vector search alpha to rag retrieval for hybrid search ranking (6624ebe)
• Adding Weaviate Vector DB option for RAG corpuses to SDK (9b28202)
• GenAI - Added system_instruction and tools support to GenerativeModel.count_tokens (50fca69)
• GenAI - Added Llama3 support in GenAI batch prediction (6166152)
• PrivateEndpoint.stream_raw_predict (197f333)
• Support reserved_ip_ranges for VPC network in Ray on Vertex cluster (36a56b9)
• Tokenization - Deprecated ComputeTokenResult.token_info_list in favor of ComputeTokenResult.tokens_info (efbcb54)

Bug Fixes

• Tensorboard - Fixed bug in tensorboard uploader where profile logs are not uploaded from nested run directories. (37627de)
• Tokenizers - Fixed Tokenizer.compute_tokens (c29fa5d)

1.64.0 (2024-08-27)

Features

• Add support for Prediction dedicated endpoint. predict/rawPredict/streamRawPredict can use dedicated DNS to access the dedicated endpoint. (3d68777)
• Add unit tests for Gen AI evaluation module (94289b5)
• GenAI - Added the model Distillation feature (private preview) (a0d4ff2)
• Grounding - Allow initialization of grounding.VertexAISearch with full resource name or data store ID, project ID, and location. (f334321)
• Make Rouge class available in base level init (aed82a1)
• Read for online store w/private service connect (7af80c6)
• Support autoscaling in Ray on Vertex (961da42)

Bug Fixes

• Fix error in tensorboard uploader thrown when time_series_id is None (d59a052)
• Fix typo in prompt templates: (c8fa7a8)
• Update generative models tests to be parameterized on environment variable name (be4cbea)

Documentation

• samples: Adding code sample for vector search create streaming index (71464e7)

1.63.0 (2024-08-20)

Features

• A new field satisfies_pzs is added to message .google.cloud.aiplatform.v1.BatchPredictionJob (#4192) (6919037)
• Add advanced PDF parsing option for RAG file import (6e1dc06)
• Add multithreading for custom metric computation. (2c93fc1)
• Add progress bar for generating inference. (b78714f)
• Add progress bar to custom metrics. (3974aec)
• Add Prompt class support for configs and Prompt.generate_content wrapper (7f1e031)
• GenAI - Added seed parameter to the GenerationConfig class (9f1e073)
• GenAI - Added the Candidate.avg_logprobs property (de80695)
• GenAI - Released the Prompt feature to Public Preview (64eeab8)
• 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)
• 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)
• GenAI Evaluation: Release GenAI Evaluation SDK GA features to vertexai.preview module. (fd38b49)
• Publish GenAI Evaluation SDK GA features to vertexai.evaluation module. Switch GenAI Evaluation Service client to v1 version. (45e4251)

Bug Fixes

• Add support of display_name to create_cached_content in python SDK (ecc2d54)
• Always upload the pickled object and dependencies tarball when creating ReasoningEngine (34ef5a3)
• Remove grouding attribution (f6ece65)

Documentation

• Update Prompt docstring for batch prompting (e96b6e6)

1.60.0 (2024-07-24)

Features

• Add preflight validations to PipelineJob submit and run methods. (c5a3535)
• Add support for langchain v0.2+ package versions in default installation (259b638)
• GenAI - Added tokenization support via GenerativeModel.compute_tokens (cfe0cc6)
• GenAI - ContextCaching - allow from_cached_content to take the cached_content resource name (8f53902)
• Make count_tokens generally-available at TextEmbeddingModel. (efb8413)

Bug Fixes

• Avoid throw error when Part.text is empty in modality content checks (bbd4a49)
• Correct logit_bias type annotation to accept keys as strings (2676d25)
• Create FV embedding dimensions sample - dimensions should be an int (2aa221e)
• Fix the sync option for Model Monitor job creation (22151e2)
• Include DeploymentResourcePool class in aiplatform top-level sdk module (ecc4f09)
• Overriding the current TracerProvider when enabling tracing (1476c10)
• Pass the project ID from vertexai.init to CloudTraceSpanExporter when enable_tracing=True for LangchainAgent (3ec043e)

Documentation

• GenAI - Update README.md for Vertex Generative AI SDK for Python to add subsections to the right nav. (42af742)

1.49.0 (2024-04-27)

Features

• Add additional parameters for GenerationConfig (0599ca1)
• Add FeatureNormType to MatchingEngineIndexConfig. (c0e7acc)
• Add mappings to pipeline templates for text-embedding models. (6279924)
• Add support for TPU v5 lite pod(v5e) for custom training jobs. Custom training jobs now accept the v5e machine types as listed in https://cloud.google.com/tpu/docs/tpus-in-gke#v5e. (415912e)
• Fix typo in Vertex Feature Store SDK. (b5404e7)
• GenAI - Improved the exception messages when candidates, parts or text are not available (e82264d)
• Support PreflightValidation in Preview PipelineJob submit function. (e88dc0d)
• Tune_model method for class TextEmbeddingModel. (3eda55d)
• Vertex RAG for enhanced generative AI (39b5149)

Bug Fixes

• GCS Bucket subdirectory not being creating while using ReasoningEngine create method (3d22a18)
• GenAI - Fixed the Part.mime_type property (819a44e)
• Handle missing import for RunnableConfig when generating schema for LangChain templates (76c5d6d)
• Parse intermediate steps from LangChain into JSON. (754c89d)
• Register TensorFlow models from Ray checkpoints for more recent TensorFlow version, addressing the deprecation of SavedModel format in keras 3 (1341e2c)
• Tensorboard uploader - move remaining tensorboard uploader code to aiplatform (f34094b)

1.48.0 (2024-04-17)

Features

• Add support for reading requirements from a file. (80db7a0)
• Adding tpu_topology to Vertex SDK (423c764)
• Enable continuous upload for profile logs. (f05924d)
• GenAI - Added the GenerationResponse.prompt_feedback property (efd5a72)
• GenAI - Added the GenerationResponse.usage_metadata property (0654c35)
• Support NOT_EQUAL for MatchingEngineIndexEndpoint numeric_restricts. (aa918e3)
• Support referenced models in SDK. (c9b6b8b)

Bug Fixes

• Add validation check for extra_packages when creating a reasoning engine. (255dabc)
• Add validation for langchain tools. (a821d50)
• Fixed the vertexai.init partial initialization issues (636a654)
• GenAI - Workaround for streaming when content role is missing in service responses (fa35b91)

Documentation

• Add Reasoning Engine reference documentation (496fc4b)
• GenAI - Add Rapid Evaluation SDK reference documentation (40b728b)

1.47.0 (2024-04-06)

Features

• Add display experiment button for tuning in Ipython environments (9bb687c)
• Add Persistent Resource ID parameter to Custom Job form_local_script, run, and submit methods. (f5be0b5)
• Add Persistent Resource Id parameter to Custom Training Job run and submit methods. (f428006)
• Added GA support for PersistentResource management (98a07dd)
• Added GA support for running Custom and Hp tuning jobs on Persistent Resources (35ecbac)
• Added the Experiment.dashboard_url property (c8eec21)
• GenAI - Added support for SafetySetting.method (probability or severity) (317ab8f)
• GenAI - Added support for supervised fine-tuning (036d2d0)
• GenAI - Added support for system instructions (4990eb6)
• GenAI - Forced function calling feature (806ef9f)
• Initial template for Langchain on Vertex. (0752a29)
• LLM - Add RLHF-tuning support for text-bison@002 (1f27c3e)
• Vertex AI Extension SDK Public Preview (137b5e1)
• Vertex AI Reasoning Engine SDK Public Preview (6aaa5d0)
• Vertex Rapid Evaluation SDK and Prompt Template for Vertex Prompt Management Public Preview (8c6ddf5)

Bug Fixes

• GenAI - Fixed response validation error during streaming (c881998)
• GenAI - Fixed the GenerativeModel's handling of tuned models from different region (bf33fb3)
• GenAI - Fixed the TuningJob dashboard URL (5367fbb)
• Reinstate persistent resource preview class and tests (765d60d)

Documentation

• Add run custom job on persistent resource sample. (53fc845)
• Add run custom job on persistent resource sample. (31100c6)

1.46.0 (2024-03-30)

Features

• GenAI - Automatic Function Calling feature (eef84c6)
• Python SDK for Vertex Feature Store. (5015d25)

1.43.0 (2024-02-29)

Features

• Add update_mask to MatchingEngineIndex upsert_datapoints() to support dynamic metadata update. (81f6a25)
• Add Ray on Vertex BigQuery read/write support for Ray 2.9 (e048e3a)
• Add read_bigquery and write_bigquery wrappers for Ray on Vertex (4739118)
• Add transport override to enable the use of REST instead of GRPC (6ab4084)
• Enable Ray cluster creation and registering TensorFlow checkpoint to Vertex with Ray version 2.9 (ff148cd)
• LLM - Add support for batch prediction to CodeGenerationModel (code-bison) (fbf2f7c)
• Support custom timeout for MatchingEngineIndex and MatchingEngineIndexEndpoint APIs. (09d1946)
• Update ray extras installation in setup.py for Ray 2.9.3 (d947304)

Bug Fixes

• Allow destination directory to exist before we copy. (0b55762)
• LVM - Fixed VideoSegmentConfig in Multimodal Embeddings API call (665e78c)

Documentation

• Update transport docstrings to include that REST is in preview (a442866)