Skip to content

CHANGELOG.md

Latest commit

 

History

History
59 lines (49 loc) · 3.08 KB

CHANGELOG.md

File metadata and controls

59 lines (49 loc) · 3.08 KB

v0.6

  • Migrated to aiogram==3.1.1, pydantic==2.4.0, if you want to keep old pydantic, use aiogram==3.0.0b7.
  • WebUI made with Vue is available! Run it with dashboard.py
  • Configuration preset switching in WebUI, to change path of default .env file, use --env on dashboard.py, additional configuration files should are stored with .env extension in env directory.
  • New configuration options: sys_webui_host, sys_api_host, sys_request_timeout, sys_api_log_level
  • Fixed TTS initialization failing when threaded mode was off
  • Fixed memory manager initialization
  • For model configuration UI the path_to_llama_cpp_weights_dir key has been added to llm_paths
  • Reply context (1-step memory) can now be toggled with llm_assistant_add_reply_context option
  • Configuration hints are available in the WebUI.
  • Model manager in the WebUI can be used to download and set up models. A few types of models are initially supported.

v0.5

  • full TTS refactoring
  • universal cross-platform os TTS provider (pyttsx4) support
  • threaded initialization (configurable via threaded_initialization option)
  • bugfixes

Breaking changes

  • tts_enable_so_vits_svc config option has been removed in favor of tts_enable_backends.
  • tts_so_vits_svc_base_tts_provider config option has been removed, you only need base voice from now on.

v0.4

  • so-vits-svc-4.1 support
  • experimental memory manager (now supports all models except LLMs with pytorch backend)

Breaking changes

  • tts_so_vits_svc_code_path config option has been renamed to tts_so_vits_svc_4_0_code_path, and tts_so_vits_svc_4_1_code_path option was added to support so-vits-svc-4.1 models, to specify that the model is a 4.1 mode, use "v": 4.1 in tts_so_vits_svc_voices.
  • llm_host has ben fixed in .env.example file, llm_ob_host has been removed

v0.3

  • Basic config state reactivity (runtime config changes in bot are reflected in .env)
  • Fixed SD model info retrieval and reinitialization
  • Global module accessibility milestone 2
  • Experimental multi-line dialog answer support
  • Speech-to-text via Whisper.cpp, Silero an Wav2Vec2
  • Seamless dialog mode voice->text->llm->voice
  • Text-to-audio via Audiocraft

v0.2

  • LLM module / provider refactoring
  • Reply chronicler with single-item history
  • Kobold.cpp and llama.cpp server support
  • SD Lora API support
  • Global module accessibility milestone 1

Breaking changes

  • MinChatGPTChronicler and GPT4AllChronicler were removed Default assistant chronicler becomes instruct, alpaca name is kept for a limited time for backwards compatibility
  • llm_ob_host has been renamed to llm_host
  • llm_active_model_type has been deprecated and replaced by two new keys: llm_backend and llm_python_model_type (when backend is pytorch)
  • In llm_python_model_type option cerebras_gpt has been renamed to auto_hf
  • path_to_cerebras_weights has been renamed to path_to_autohf_weights
  • sd_available_loras has been deprecated, lora API endpoint is used

v0.1

  • initial version with incremental changes and full backwards compatibility with previous commits