LLMMaster Changelog

[Unreleased]

Text-To-Image models pending to be added
- Midjourney API through ImaginePro
- Adobe Firefly API not released for personal use
- Ideogram API not released for personal use
- Leonardo AI API pending to be added
Text-To-Music models pending to be added
- Mubert API for rather enterprise use
- Suno API for rather enterprise use
Pending additional functions
- OpenAI and Stable Diffusion Image-To-Image for online file.
- ElevenLabs using native REST API like dubbing and cloning, etc.
- Pika.art video extension, etc.
- OpenAi structured output
- Anthropic Claude cache
- SambaNova AI API
- Google VertexAI

[0.8.0] - 2024-10-13

Added

Skybox models
- SkyboxTextToPanorama
- SkyboxPanoramaToImageVideo (export function)

[0.7.0] - 2024-10-06

Added

Runway video models
- RunwayImageToVideo

[0.6.0] - 2024-09-27

Added

Tripo 3D modeling APIs
- TripoTextTo3D
- TripoImageTo3D
- TripoMultiviewTo3D
- TripoRefineModel
- TripoAnimationPreRigCheck
- TripoAnimationRig
- TripoAnimationRetarget
- TripoStylization
- TripoConversion

Changed

Changed output type to dict for some models, see RESULTSTYPE.md for details
Fully revised the pytest test functions
Requirement of Python version changed to 3.10 or later, 3.9 or earlier is no longer supported

[0.5.1] - 2024-09-20

Added

Flux.1 image models via fal-ai
- Flux1FalImageToImage
- Flux1FalTextToImage
ElevenLabs audio models
- ElevenLabsTextToSoundEffect
- ElevenLabsAudioIsolation

Changed

Commonized functions from BaseModel
- _wait(): waiting response from API provider
- _sanitize_url(): sanitizing URL text

[0.5.0] - 2024-09-17

Added

Luma Dream Machine video generation models
- LumaDreamMachineTextToVideo
- LumaDreamMachineImageToVideo
- LumaDreamMachineVideoToVideo

[0.4.1] - 2024-09-06

Added

CerebrasLLM class
MistralLLM class

[0.4.0] - 2024-09-04

Added

self.api_key_pairs and set_api_keys() added to LLMMaster class
_load_api_key() added to LLMInstanceCreator class

Changed

Revised all the models to set API key from either os.getenv() or self.api_key_pairs
Changed ACTIVE_MODELS in llmmaster.py from list to dictionary

[0.3.4] - 2024-08-13

Changed

SUMMON_LIMIT changed to 100 as default
Deleted wait_for_starting parameter verification from LLMMaster class

Added

summon_limit parameter added to LLMMaster class

[0.3.3] - 2024-08-12

Added

Text-To-Speech model class
- VoicevoxTextToSpeech

[0.3.2] - 2024-08-07

Changed

WAIT_FOR_SUMMONING changed to WAIT_FOR_STARTING
For Meshy models, changed to stop polling result when status is SUCCEEDED or FAILED or EXPIRED

[0.3.1] - 2024-08-04

Changed

For Meshy models, changed to stop polling result when status is SUCCEEDED or FAILED

[0.3.0] - 2024-08-01

Added

Pika.art third-party Text-To-Video model class
- PikaPikaPikaGeneration

[0.2.3] - 2024-07-30

Added

ElevenLabsTextToSpeech

[0.2.2] - 2024-07-29

Added

support Meshy 3D modeling APIs
- MeshyTextToTexture
- MeshyTextTo3D
- MeshyTextTo3DRefine
- MeshyTextToVoxel
- MeshyImageTo3D

[0.2.1] - 2024-07-28

Changed

return value changed to requests.models.Response class from response.content for the following models:
- StableDiffusionTextToImage
- StableDiffusionImageToVideo
- StableDiffusionImageToImage

Added

RESULTSTYPE.md for brief description how to handle generated contents

[0.2.0] - 2024-07-27

Added

Speech-To-Text model class
- GoogleSpeechToText

[0.1.7] - 2024-07-21

Added

Image-To-Video model class
- StableDiffusionImageToVideo

Changed

Renamed OpenAIAudioToText to OpenAISpeechToText
Renamed INSTANCE_CLASSES to ACTIVE_MODELS in llmmaster.py
Changed MAX_TOKENS to DEFAULT_TOKENS in config.py
LLMMaster no longer checks max_tokens and sets default if not given properly

[0.1.6] - 2024-07-19

Added

Image-To-Image model classes
- OpenAIImageToImage
- StableDiffusionImageToImage
elapsed_time attribute to LLMMaster class

Changed

adding dummpy prompt process to LLMInstanceCreator.verify()
OpenAITextToImage returns ImagesResponses class, changed from plain text of image URL

[0.1.5] - 2024-07-18

Added

Audio-To-Text model class
- OpenAIAudioToText

[0.1.4] - 2024-07-17

Added

Video-To-Text model class
- GoogleVideoToText

[0.1.3] - 2024-07-16

Added

Image-To-Text model classes
- OpenAIImageToText
- GoogleImageToText

[0.1.2] - 2024-07-14

Added

Text-To-Image model classes
- OpenAITextToImage
- StableDiffusionTextToImage

[0.1.1] - 2024-07-12

Changed

renamed package name from llm-master to llmmaster because hyphens and underscores were mixed up and made confusion

Removed

ver. 0.1.0 from llm-master and start 0.1.1 for new PyPI repository llmmaster

[0.1.0] - 2024-07-10

Added

BaseModel class in base_model.py
verify() function to LLMInstanceCreator class
new folders, llm_master and tests
test functions using pytest in folder tests and tested
initial PyPI uploaded

[0.0.4] - 2024-07-09

Added

Text-To_Image model
- OpenAITTI
config.py
setup.py

Changed

separated files text_to_text_models.py, text_to_image_models.py

[0.0.3] - 2024-07-09

Changed

LLMInstanceCreator.create(), simplified the process of creating LLM instances

[0.0.2] - 2024-07-08

Added

LLMInstanceCreator class in llm_master.py for creating and verifying LLM instances
LLMMaster.pack_parameters() method to help create entry dictionaries
LLMMaster.dismiss() method to clear instances and results
set limit to 32 for max instances
run each thread in each 1 sec of interval due to API provider limitation

Removed

LLMMaster.summon_all() method
LLMMaster.split_llm_information() method

[0.0.1] - 2024-07-05

Added

uploaded the first version to GitHub