{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreigupr5uetijyw67dqs7n6rzq3kgxfilbtlh3goh4a5ixpzf557sgm",
    "uri": "at://did:plc:pgryn3ephfd2xgft23qokfzt/app.bsky.feed.post/3miys7yageju2"
  },
  "path": "/t/duplicated-space-not-working-as-original/175072#post_2",
  "publishedAt": "2026-04-08T12:23:58.000Z",
  "site": "https://discuss.huggingface.co",
  "tags": [
    "Ratnesh-dev/parakeet-tdt-0.6b-v3 · Fix version drift",
    "[NeMo 2.6] RNNT ASR inference fails on A100 (CUDA 12.8, PyTorch 2.9) with CUDA Graphs error CUDA failure! 35 · Issue #15145 · NVIDIA-NeMo/NeMo · GitHub",
    "TDT Inference uses a lot of CPU and slow down with `use_cuda_graph_decoder: True` · Issue #15164 · NVIDIA-NeMo/NeMo · GitHub"
  ],
  "textContent": "After cloning the repository and pinning the dependencies to older versions, transcription now completes successfully. So I’ve sent PR.\n\nThe cause was likely that PyTorch on Zero GPU defaulted to version 2.9, combined with the fact that PyTorch 2.9 and the new NeMo don’t seem to play well together.\nAlso, I’m still not quite sure how to stabilize this Space with the new PyTorch and NeMo…\n\n### PR\n\n  * Ratnesh-dev/parakeet-tdt-0.6b-v3 · Fix version drift\n\n\n\n### Related issue\n\n  1. [NeMo 2.6] RNNT ASR inference fails on A100 (CUDA 12.8, PyTorch 2.9) with CUDA Graphs error CUDA failure! 35 · Issue #15145 · NVIDIA-NeMo/NeMo · GitHub\n  2. TDT Inference uses a lot of CPU and slow down with `use_cuda_graph_decoder: True` · Issue #15164 · NVIDIA-NeMo/NeMo · GitHub\n\n\n\nEdit:\nYou maybe need to do `Factory rebuild` to fix the Space after merged.",
  "title": "Duplicated Space Not Working As Original"
}