Fix Gradio Progress IndexError by using is not None

Fix noisereduce, UI progress, remove spaces.GPU for L4, and add train_cli.py

Optimize RVC training: resumable training with checkpoints every 10 epochs and advanced UI options

Fix RVC training module structure and imports

Add __init__.py files and fix sys.path for rvc_logic

Fix build error: use local rvc_logic (cleaned)

Migrate RVC training to ultimate-rvc library (no fairseq needed)

Integrate RVC training tab (100 epochs)

feat: add robust logging and diagnostics for ZeroGPU troubleshooting; fix '0 seconds' issue with explicit checks

fix: move all output and storage directories from /tmp to local app subdirectories to fix Gradio serving issues on HF; explicitly allow results/checkpoints paths; set Audio type to filepath

fix: lazy-import torch/torchaudio in inference.py to avoid libcudart.so.13 crash; pin cu124 wheels in requirements to prevent demucs pulling CUDA-13 torchaudio

fix: consolidate Demucs+SeedVC+Mix into single @spaces.GPU function to fix ZeroGPU nested GPU call error; expose _separate_audio_impl; add libsox-dev to packages.txt

fix: add gradio==5.12.0 to requirements, fix de-esser bug (LowShelf->HighShelf), show_error=True

Add voice similarity control + improve reference audio processing

Upgrade audio quality: pro mixing chain, better inference params, htdemucs_ft

Fix GPU timeout: increase to 300s, lower default diffusion steps to 10

Fix inference tensor backward error: add torch.no_grad + inference_mode

Add GPU worker debug logging to diagnose ZeroGPU import errors

Fix missing 'munch' dep + add detailed error wrapper for ZeroGPU

Bundle Seed-VC modules directly in repo for ZeroGPU compatibility

Fix ModuleNotFoundError: clone Seed-VC to app dir instead of /tmp

Replace RVC with Seed-VC for zero-shot voice conversion

Implement real RVC v2 inference pipeline with HuBERT + FAISS + generator

Convert pretrained checkpoint to RVC inference format for proper voice conversion

Upgrade to Gradio 5.12.0 for proper ZeroGPU authentication

Remove @spaces.GPU from separation and inference for CPU fallback

Rewrite inference with fallback: try Applio, then pitch-shift + FAISS

Skip HiFi-GAN training entirely, use pre-trained model + FAISS index

Better error messages + reduce segments to 40 for GPU time limit

Restore full quality params now that GPU is available

Reduce all @spaces.GPU durations to 60s (ZeroGPU max limit)

Skip HiFi-GAN training on CPU, use pre-trained model + FAISS index

Optimize training for CPU: limit slices, reduce batch size and epochs

Fix 'context has already been set' multiprocessing error

Replace Applio preprocess subprocess with custom implementation

Search for sliced audio recursively and show directory tree in errors

Add detailed logging to preprocessing to debug missing audio slices

Initial commit: Clone Vocal RVC - web voice cloning tool