* update script
* update headers
* refactor a bit and add test script
* update headers
* update for edge case
* update headers
* add step to get file creation date
* use git history to get copyright year
* generation type is printed with inference parameters
* fix unit test
* Add generation type to ModelConfig
* pass tests
* added generate_text_embeddings
* tests
* remove sensitive=True old artifact no longer needed
* Slight refactor
* slight refactor
* Added embedding generator
* chunk_separator -> chunk_pattern
* update tests
* rename for consistency
* Restructure InferenceParameters -> CompletionInferenceParameters, BaseInferenceParameters, EmbeddingInferenceParameters
* Remove purpose from consolidated kwargs
* WithModelConfiguration.inference_parameters should should be typed with BaseInferenceParameters
* Type as WithModelGeneration
* Add image generation modality
* update return type for generate_kwargs
* make generation_type a field of ModelConfig as opposed to a prop resolved based on the type of InferenceParameters
* remove regex based chunking from embedding generator
* Remove image generation for now
* more tests and updates
* column_type_is_llm_generated -> column_type_is_model_generated
* change set to list: fix flaky tests
* CompletionInferenceParameters -> ChatCompletionInferenceParameters for consistency with generation_type
* Update docs
* fix deprecation warning originating from cli model settings
* update display of inference parameters in cli list
* save prog on inference parameter
* updates for the ocnfig builder
* update cli readme
* update cli for inference parmeters
* update inference parameter names
* flip order of vars
* WithCompletion -> WithChatCompletion
* specify InferenceParamsT
* Update columns.md with EmbeddingColumnConfig info
* make generation_type a descriminator field in inference params. add configuration support for max_parallel_requests and timeout
* DRY out some stuff in field.py
* Update nomenclature. prompt tokens -> input tokens, completion tokens -> output tokens in column statistics for consistency
* Add nvidia-embedding and openai-embedding to default model configs
* Fix typo in docs
* Make generate collab notebooks
* fine-tune -> adjust
* column config -> column name when possible
* fallback to dtype of first non-null element
* add unit tests
* add error message info to warning
* catch str_ too
* separate column configs and types
* create plugin object
* create plugin manager
* fix config integration
* make base task registry raise on collision false by default
* update registry test after raise on collision default update
* make analysis work using general stats calculation
* default -> builtin
* use entry point approach instead
* rewire using plugin helpers
* add env var to disable plugins
* fix tests
* update plugin manager tests
* add tests for plugin helpers
* update license headers
* add emoji
* not using the pm in the builder code
* Update src/data_designer/plugins/manager.py
Co-authored-by: Nabin Mulepati <nmulepati@nvidia.com>
* Update src/data_designer/plugins/manager.py
Co-authored-by: Nabin Mulepati <nmulepati@nvidia.com>
* Update src/data_designer/plugins/manager.py
Co-authored-by: Nabin Mulepati <nmulepati@nvidia.com>
* merge plugin registry into the manager
* small pr feedback
* client side plugin manager
* builtin -> default; move adding plugins to registry
* update method names to better match what they do
* use register verb for consistency with other registries
* thread safety updates; make discover private
---------
Co-authored-by: Nabin Mulepati <nmulepati@nvidia.com>
