Tags

Branch force pushed by vivekp27 at 2024-08-29T12:42:51+00:00, 3470652-chunked-embeddings-do moved from this tag to 7963df85b743e402e8410471d82597e523467d44

Branch force pushed by vivekp27 at 2024-08-29T12:34:47+00:00, 3470652-chunked-embeddings-do moved from this tag to d341de707eab043bf2d8a8ebeced1827e2416ee2

Branch force pushed by vivekp27 at 2024-08-29T12:33:26+00:00, 3470652-chunked-embeddings-do moved from this tag to 430be465f65c804fe54874875dc12ecee39e6836

Branch force pushed by vivekp27 at 2024-08-29T12:31:49+00:00, 3470652-chunked-embeddings-do moved from this tag to b576722d3ef921688bda483959b52eb8fc17cbec

Fix for AI Assistants render service

* First AI Search
* Only store the highest score of an result for entities
* Mixed message on vectorSearch and querySearch on thjeAiVdbProviderInterface
* Make more embedding strategy options configurable
* Refactor embedding strategies to be able to extend easier
* Make Vector Search Method Configurable?
* Remove and trim the output in the AI Translate module
* Create a generic plugin layer for CKEditor
* Add access check and iterator logic into SearchApiAiSearchBackend
* Listing pages of provider and AI Logging doesn't show the right listing
* Prompt output doesn't work in logging
* Add new CKEditor plugins to the drop down button in CKEditor toolbar
* Create Entity Tab with visual Automators for an entity
* Delete backend should delete Milvus collection
* Small Zilliz Bugs
* Adding host with slash in the end breaks everything
* Document how to run tests in Milvus dependency
* Error when adding AI Assistant without default provider set in AI Settings
* Add warning message to configure default operation type settings.
* Add access check and iterator logic into SearchApiAiSearchBackend

Release notes
* First Version of AI Automator
* Embeddings should allow to specify the number of dimensions
* Image to Video Operation Type
* Move the ECA processor into the project and move up to ECA 2.0
* Code suggestion for text-to-image is old in the explorer
* Base64 response does not work on OpenAI with normalizers on text-to-image
* Refactor all files to use generictype files
* Garbage Cleaning Configs
* Add Speech-to-Speech explorer
* Add Audio to Audio Explorer
* Remove stream parameter from normal chat and add streamChat operation type.
* Todo: Audio to Audio operator type
* Add Speech-to-Speech operation type
* Port openai_content to ai module
* Generic AI CKEditor Plugin
* ai translate (ai_translate) submodule
* ai_validations for smart text and image validation
* LLM modules ajax fails when Token module is missing.
* Switch Custom Field descriptions fields to textarea
* Remove Install warning message about missing automators
* Fix phpcs issues before alpha4
* Set up experimental modules
* Discussion: Streaming Input/Output
* Move to embeddings to ai provider and allow to select the embeddings model when embedding a field. Move to embeddings to ai provider and allow to select the embeddings model when embedding a field.

New Providers

New design and initial bugs

Initial release of the AI module!