Upload ui.py with huggingface_hub

Fix: add all TTS parameters (text_split_method, seed, etc.) for proper long text handling

Fix: correct reference text for zh_vo_MAIN_YHX_2_12.wav

Fix: use correct reference text matching zh_vo_MAIN_YHX_2_12.wav

Fix: update ref text and prompt lang to Chinese for Aemeath

Fix: use Aemeath reference audio (zh_vo_MAIN_YHX_2_12.wav)

Fix: add uvicorn.run() to actually start the server

Revert: remove COPY model files, use huggingface_hub download instead

Fix: use huggingface_hub to download models (xet compatible)

Fix: COPY model files into image instead of downloading at runtime

Upload reference audio for Aemeath

Upload aimisi-e16.ckpt

Upload README.md with huggingface_hub

Upload 2 files

Delete 爱弥斯.zip with huggingface_hub

Upload start.sh with huggingface_hub

Upload start.sh with huggingface_hub

Upload 爱弥斯.zip

Download required NLTK data at startup

Add GSVI compatibility routes

Preload fast-langdetect model

Create fast-langdetect cache directory

Fix Space boot and wire up TTS UI

v4: move model downloads to runtime, simplify build

v4: move model downloads to runtime, simplify build

Fix v2: proper build-essential, minimal deps, CPU config

Fix v2: proper build-essential, minimal deps, CPU config

Fix: CPU-only Dockerfile with pretrained models, proper start.sh

Fix: CPU-only Dockerfile with pretrained models, proper start.sh

Fix: CPU-only Dockerfile with pretrained models, proper start.sh

wget tarball instead of git clone

pure CPU build, no CUDA deps

pure CPU build, no CUDA deps

Fix: add fallback torchaudio reinstall, tts_infer.yaml creation

Fix: force reinstall torchaudio for CUDA 12.6 at build time using conda pip

Fix: remove invalid -d cpu flag, add tts_infer.yaml fallback, better logging

Fix: reinstall torchaudio for CUDA 12.6 to match PyTorch

Update start.sh for official image with model symlinks

Use xxxxrt666/gpt-sovits:latest-cu126-lite official image

use official gpt-sovits docker image

use official gpt-sovits docker image

fix: sequential download, let api_v2 handle hubert/bert

fix: add gradio dep

connect api_v2.py inference

connect api_v2.py inference

auto download models on startup

auto download models on startup

switch to FastAPI

switch to FastAPI

switch to FastAPI