James Pine 6bf40bd2d0 fix tokenizer patch corrupting AutoTokenizer for other engines ...

Replace the monkey-patch on AutoTokenizer.from_pretrained (which broke
the classmethod descriptor and caused 'Tokenizer not loaded' errors
when loading Qwen after TADA) with two targeted config patches:
- Set AlignerConfig.tokenizer_name to the local ungated tokenizer path
- Pre-load TadaConfig, inject tokenizer_name, pass config= to from_pretrained

No global state is modified; other engines are unaffected.

2026-03-17 03:15:57 -07:00

..

__init__.py

add HumeAI TADA TTS engine (1B English + 3B Multilingual)

2026-03-17 01:55:15 -07:00

base.py

comment cleanup

2026-03-16 01:46:19 -07:00

chatterbox_backend.py

refactor: remove dead code, deduplicate backends

2026-03-16 01:10:02 -07:00

chatterbox_turbo_backend.py

refactor: remove dead code, deduplicate backends

2026-03-16 01:10:02 -07:00

hume_backend.py

fix tokenizer patch corrupting AutoTokenizer for other engines

2026-03-17 03:15:57 -07:00

luxtts_backend.py

refactor: remove dead code, deduplicate backends

2026-03-16 01:10:02 -07:00

mlx_backend.py

add model selection and expanded language support to /transcribe endpoint

2026-03-16 22:44:28 -07:00

pytorch_backend.py

add model selection and expanded language support to /transcribe endpoint

2026-03-16 22:44:28 -07:00