vachan-ai upgrade (25.5.28-beta)

[NoelSudhish]

released new image vachan-ai:25.5.28-beta
Release notes

New features:

Model logging feature for fine-tuned NLLB models
Voice cloning with ToucanTTS (pre-trained and fine-tuned models)
    Included new param in voice clone api - transcripts of input audios (should be a csv file having columns: audio, transcripts)
PUT API to edit language in the database
DELETE API to delete language from the database
Speaker references included in the following apis
    TTS Doc api, Voice clone
    user can select any available speaker reference or can upload new reference audio
TTS Doc api: audios generated for multiple input text can be merged into a single audio by setting merge_audios flag as true

Enhancements:

Improved model dataclass structure
Detailed logs for better analysis of process latency
GET language api accepts lang_id also
Enhanced model serving: utils zip not required in the serving side

Docs

Updated swagger docs
Updated README
Updated models csv - details of api dependencies added for each model

Code refactoring:

Split inference core (having core functions of different services) into small modules for following features
    Machine translation
    Text-to-speech