Commit History

  1. add 2

    7ab065c
  2. sync local state with remote

    Verified signature
    1c7b2ff
  3. remove usage instructions and example text from UI

    Verified signature
    2af14be
  4. fix: use torchaudio.info with suppressed deprecation warnings instead of torchcodec

    Verified signature
    9e44af9
  5. fix: convert non-WAV audio to WAV before encoding, use torchcodec for duration

    Verified signature
    9508f19
  6. fix: use torchaudio.info for audio duration (supports m4a without deprecation warnings)

    Verified signature
    3a6a578
  7. fix: use librosa.get_duration for audio duration check (supports m4a and more formats)

    Verified signature
    d99970c
  8. fix: only unpatch ZeroGPU in nanovllm child processes, skip @spaces.GPU workers

    Verified signature
    0443284
  9. fix: forcibly clear ZeroGPU dispatch/function mode stacks in child processes

    Verified signature
    3e0652f
  10. fix: replace ZeroGPU dispatch stubs with pass-through in child processes instead of trying to pop modes

    Verified signature
    41ea91f
  11. fix: properly unpatch ZeroGPU dispatch modes and force torch.load map_location=cpu in child processes

    Verified signature
    d0f587a
  12. fix: exclude assets/ and examples/ from LFS tracking

    Verified signature
    99506e9
  13. fix: use functools.partial for pickleable multiprocessing wrapper

    71f1902
  14. fix: make spaces torch mode __exit__ no-op to avoid PyTorch 2.8.0 IndexError

    dd79864
  15. fix: unpatch ZeroGPU torch stubs in every multiprocessing child process

    033d78c
  16. fix: import os before using os.environ in top-level patches

    06f5b3b
  17. fix: patch spaces ZeroGPU workers to non-daemon to allow child processes

    2ca879f
  18. fix: use direct env check to disable prewarm on ZeroGPU

    9f85f57
  19. fix: wrap TTS inference in @spaces.GPU for ZeroGPU compatibility

    f170d0c
  20. fix: suppress harmless asyncio BaseEventLoop __del__ errors during GC

    619c24e
  21. fix: close asyncio event loop in worker thread to suppress __del__ errors

    2da9298
  22. fix: upgrade torch to 2.8.0 for ZeroGPU compatibility

    39a0ffd
  23. fix: import spaces before other packages to avoid CUDA init error

    6757fca

  1. fix: harden async nanovllm demo stability⁠

    εˆ˜ι‘«
    44d9bcd
  2. fix: revert concurrency limit to 1 β€” nanovllm event loop is not thread-safe⁠

    εˆ˜ι‘«
    364d86e

  1. fix: remove unsupported inference_timesteps parameter and dit_steps slider⁠

    εˆ˜ι‘«
    0203b68

  1. fix: add cookbook guide heading link⁠

    εˆ˜ι‘«
    5351045
  2. fix: store daily request logs on persistent disk⁠

    εˆ˜ι‘«
    39d1990
  3. fix: show localized upload alert for long audio⁠

    εˆ˜ι‘«
    1cf9f24
  4. fix: zipenhancer for denoise

    εˆ˜ι‘«
    96f588d
  5. fix: add denoiser and reference audio guards⁠

    εˆ˜ι‘«
    ffbf382

  1. fix: harden Gradio NanoVLLM startup path⁠

    εˆ˜ι‘«
    7d8de04
  2. fix: install flash-attn from prebuilt wheel⁠

    εˆ˜ι‘«
    d138332
  3. fix: pin Gradio NanoVLLM runtime dependencies⁠

    εˆ˜ι‘«
    ed56d7a
  4. fix: switch Gradio demo backend to nanovllm_voxcpm⁠

    εˆ˜ι‘«
    09dc185
  5. add torch version && torch compile == true

    εˆ˜ι‘«
    29b439e
  6. add torch version

    εˆ˜ι‘«
    9dc8c20

  1. fix: dynamically install torchaudio matching ZeroGPU torch+CUDA version⁠

    εˆ˜ι‘«
    d3b2b16
  2. fix: install CPU torchaudio to avoid CUDA 13 mismatch on ZeroGPU⁠

    εˆ˜ι‘«
    f15ea05
  3. fix: install voxcpm with --no-deps at runtime to preserve ZeroGPU torch/torchaudio⁠

    εˆ˜ι‘«
    59f3c75
  4. fix: remove torch/torchaudio pin, rely on ZeroGPU pre-installed version⁠

    εˆ˜ι‘«
    60d19ef
  5. fix: pin torch/torchaudio to 2.5.0 and move import voxcpm to module level⁠

    εˆ˜ι‘«
    1253e36
  6. fix: move SDPA backend config into GPU function context⁠

    εˆ˜ι‘«
    b84906b
  7. fix: disable flash/mem-efficient SDPA backends for ZeroGPU compat⁠

    εˆ˜ι‘«
    8f661ec
  8. fix: disable optimize to avoid ZeroGPU torch.compile incompatibility⁠

    εˆ˜ι‘«
    d45dc51
  9. upgrade to VoxCPM2 + Gradio 6 with full i18n UI⁠

    εˆ˜ι‘«
    7d9f729

  1. Update README.md

    Verified signature
    ced8c00

  1. set zero gpu inference

    εˆ˜ι‘«
    f6f653d
  2. set zero gpu inference

    εˆ˜ι‘«
    1700cda
  3. set zero gpu inference

    εˆ˜ι‘«
    1488551