Skip to content

Latest commit

 

History

History
114 lines (88 loc) · 6.6 KB

CHANGELOG.md

File metadata and controls

114 lines (88 loc) · 6.6 KB
Raw

. minor | - Fix | + Addition | ^ improvement | ! Change | * Refactor

IMPORTANT: 0.1.x will still have some breaking changes in patches.

  • Make sure to lock your version, e.g., genai = "=0.1.9".
  • Version 0.2.x will follow semver more strictly.
  • API changes will be denoted as "! - API CHANGE ...."

2024-09-18 - 0.1.9

  • ^ AdapterKind - expose default_key_env_name
  • . openai - add 'o1-' model prefix to point to OpenAI Adapter
  • . comments proofing (using genai with custom devai script)
  • . #23 - add documentation
  • . fix printer comment
  • . updated to v0.1.9-wip

2024-09-06 - 0.1.8

  • . printer - now uses printer::Error (rather than box dyn) (rel #21)
  • + NEW - structured output - for gemini & OpenAI
    • Behind the scene:
      • Static Badge Adding Gemini Structured Output (vid-0060)
      • Static Badge Adding OpenAI Structured Output (vid-0059)
  • ! soft deprecation (for now) use ChatResponseFormat::JsonMode (was ChatOptions::json_mode flag)
  • * Make most public types De/Serializable
  • . openai - fix chatgpt prefix. Update current model lists
  • . add json test for anthropic
  • . makes webc::Error public (relates to: #12)

2024-08-14 - 0.1.7

  • + Added ModelMapper scheme (client_builder::with_model_mapper_fn)
    • Static Badge - genai ModelMapper code demo (v0.1.7)
  • ! API CHANGE Removed AdapterKindResolver (should use ModelMapper) (see examples/c03-mapper.rs)
  • ! API CHANGE Renamed ModelInfo to ModelIden
  • ! API CHANGE AuthResolver - Refactor AuthResolver Scheme/API (see examples/c02-auth.rs)
  • ! API CHANGE completely remove AdapterConfig (see AuthResolver)
  • . test groq - switch to llama3-groq-8b-8192-tool-use-preview for testing to have the test_chat_json work as expected
  • ^ chore: make stream is send
  • . test - ChatOptions - add tests for temperature
  • . A typo in adapters for OpenAI makes the temperature chat option unusable.
  • . unit test - first value_ext insert

2024-07-26 - 0.1.6

  • + ChatOption Add json mode for openai type models
  • . groq - added the Llama 3.1 previews, and grog-..-tool-use.. to the groq model list names
  • ! now chat::printer::print_chat_stream (was utils::print_chat_stream)
  • ! Now ChatOptions (was ChatRequestOptions) ! Remove client_builder.with_default_chat_request_options (available with client_builder.with_chat_options)
  • . readme - add youtube videos doc

2024-07-21 - 0.1.5

  • ! API CHANGE now ClientBuilder::insert_adapter_config (was with_adapter_config)
  • . code clean

2024-07-19 - 0.1.4

  • ! API CHANGE - refactor Error
    • With new ModelInfo
    • Back to genai::Error (adapter::Error was wrongly exposing internal responsibility)
  • . update tests and examples from 'gpt-3.5-turbo' to 'gpt-4o-mini'
  • - Fix naming ClientConfig::with_adapter_kind_resolver (was wrongly ...auth_resolver)
  • * refactor code layout, internal Adapter calls to use ModelInfo
  • + Add ModelName and ModelInfo types for better efficient request/error context
  • ! API CHANGE - now Client::resolve_model_info(model) (was Client::resolve_adapter_kind(mode))
  • ^ ChatRequest - add ChatRequest::from_system
  • . updated provider supported list

2024-07-18 - 0.1.3

  • ^ openai - added gpt-4o-mini and switched all openai examples/tests to it
  • ! API CHANGE - New MessageContent type for ChatMessage.content, ChatResponse.content, and StreamEnd.captured_content (only ::Text variant for now).
    • This is in preparation for multimodal support
  • ! API CHANGE - (should be minor, as Into implemented) - ChatMessage now takes MessageContent with only ::Text(String) variant for now.
  • ! API CHANGE - Error refactor - added genai::adapter::Error and genai::resolver::Error, and updated genai::Error with appropriate Froms
  • + Added token usage for ALL adapters/providers - ChatResponse.usage and ChatRequestOption .capture_usage/.capture_content (for streaming) support for all Adapters (see note in Readme for Ollama for streaming)
  • ! API CHANGE: ClientConfig::with_chat_request_options (was with_default_chat_request_options)
  • ! API CHANGE: PrintChatStreamOptions::from_print_events (was from_stream_events)
  • ^ AdapterKind - added as_str and as_lower_str
  • ^ ChatRequest - added .iter_systems() and .combine_systems() (includes eventual chat_req.system as part of the system messages)
  • ! API CHANGE: Client::all_model_names(..) (was Client::list_models(..))
  • ^ groq - add gemma2-9b-it to the list of Groq models
  • ! API CHANGE: genai::Client (was genai::client::Client, same for ClientBuilder ClientConfig)
  • - groq - remove groq whisper model from list_models as it is not a chat completion model
  • ^ ollama - implement live list_models for ollama
  • ! Makes AdapterDispatcher crate only (should be internal only)

2024-07-08 - 0.1.2

  • + ChatRequestOptions - added temperature, max_tokens, top_p for all adapters (see readme for property mapping).
  • ! SyncAdapterKindResolverFn - Change signature to return Result<Option> (rather than Result)
  • . made public client.resolve_adapter_kind(model)
  • + implement groq completions

2024-06-12 - 0.1.1

  • - gemini - proper stream message error handling

2024-06-11 - 0.1.0

  • . print_chat_stream - minor refactor to ensure flush

2024-06-10 - 0.0.14

  • - ollama - improve Ollama Adapter to support multi system messages
  • - gemini - fix adapter to set "systemInstruction" (Supported in v1beta)

2024-06-10 - 0.0.13

  • + Added AdapterKindResolver
  • - Adapter::list_models api impl and change
  • ^ chat_printer - added PrintChatStreamOptions with print_events