Spaces:
Runtime error
Runtime error
Commit History
-
refactor based on DDD
Michael Hu5009cb8 -
fix circular dependency
Michael Huaaa0814 -
remove fallback to Dummy TTS
Michael Hue22e786 -
replace dummy tts class
Michael Huae641cf -
fix path
Michael Hub2b15db
-
refactor tts module
Michael Hu7495571 -
add faster whisper installation
Michael Hu37aaac6
-
refactor(stt): replace whisper with faster-whisper for improved performance
Michael HuSwitch from transformers-based whisper implementation to faster-whisper for better speed and memory efficiency. The new implementation removes torch dependency for device detection and uses optimized compute types based on available hardware.b591083
-
use unsloth whisper model to reduce the model size
Michael Hu6f92dbc
-
set correct fallback sequence
Michael Hu6034fea -
create fallback flow for tts engines
Michael Hu58d9769 -
fix
Michael Hue734196 -
fix import issue
Michael Hucb90410 -
handle dia model not available
Michael Hua316f58 -
refactor tts
Michael Hu60bd17d -
default to parakeet model
Michael Hu27972f7
-
refator tts part
Michael Hu3ed3b5a -
default to parakeet model as it is much faster
Michael Hu31708ca -
support nvdia Parakeet TDT ASR
Michael Hu0ee4f42
-
fix
Michael Hue68cae8 -
fix gradio too many requests exception
Michael Hu451d2b1 -
handle voice generation failure exception
Michael Hufeebf14 -
update to set a voice
Michael Hue3843d2 -
use kokoro fastAPI server to generate voice
Michael Hu22bd0b9 -
try to use kokoro api server as backup
Michael Hu5a72681
-
upgrade streamlit version
Michael Hu724aa35 -
re-activate dia
Michael Hud5c9fdf -
temporarily disable dia
Michael Hubb33c13 -
downgrade streamlit to ensure compatiblity with descript-audio-codec
Michael Hu086a552 -
try fixing build issue
Michael Hub697102 -
revert to earlier version of kokoro because of a incompatiblity issue
Michael Huad6b877 -
Revert "Update README.md"
Michael HuThis reverts commit 5b2712563bbfa23c72e8ae22c408a748ad20238c.c549dab -
Update README.md
Verified signature5b27125 -
Switch to use Gradio
Verified signaturee2894f6 -
fix box is not a valid method in gradio issue
Michael Hu6514e96 -
move to Gradio so we can leverage ZeroGPU
Michael Huf0248ed -
fix runtime dia audio is not generated issue
Michael Hu4a9bb1a -
fix runtime dia model issue
Michael Hu2d176f4 -
try resolve buld issue
Michael Hu501f0b5 -
try resolve build issue
Michael Hu44fb7ee -
enhance logging
Michael Hu91223c9 -
add missing dac (descript-audio-codec) import
Michael Hu419e343 -
use dia tts as fallback model if kokoro is not available
Michael Hu7b25fdd -
add dia tts model
Michael Hu030c851 -
add sample dia app file
Michael Hu4b0381b -
add gitignore file
Michael Hu875a169 -
fix init file
Michael Hu28a10f3
-
add dia tts model. Since dia is not yet released to pypi, we pull in the source directly
Michael Hu9c4b958
-
chore: add espeak-ng to packages.txt
Michael HuThis commit adds the espeak-ng package to the packages.txt file for dependency management.1a3633a -
build: add phonemizer-fork dependency to requirements.txt
Michael HuAdd phonemizer-fork>=3.3.2 to the project dependencies to support new features requiring phonemization functionality.672e7e1