From 5afa2fe656a49706bdc20cb778fd04e66dac7aec Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Alexandre=20D=C3=A9fossez?= Date: Thu, 3 Jul 2025 15:06:28 +0200 Subject: [PATCH] plop --- scripts/tts_pytorch.py | 3 +-- tts_pytorch.ipynb | 4 +--- 2 files changed, 2 insertions(+), 5 deletions(-) diff --git a/scripts/tts_pytorch.py b/scripts/tts_pytorch.py index 3e36613..f513aaa 100644 --- a/scripts/tts_pytorch.py +++ b/scripts/tts_pytorch.py @@ -65,7 +65,7 @@ def main(): voice_path = tts_model.get_voice_path(args.voice) # CFG coef goes here because the model was trained with CFG distillation, # so it's not _actually_ doing CFG at inference time. - # Also, if you are generating a dialog, you should have at least two voices in the list. + # Also, if you are generating a dialog, you should have two voices in the list. condition_attributes = tts_model.make_condition_attributes( [voice_path], cfg_coef=2.0 ) @@ -76,7 +76,6 @@ def main(): pcms = queue.Queue() - @torch.no_grad() def _on_frame(frame): if (frame != -1).all(): pcm = tts_model.mimi.decode(frame[:, 1:, :]).cpu().numpy() diff --git a/tts_pytorch.ipynb b/tts_pytorch.ipynb index ab6e228..9680bdd 100644 --- a/tts_pytorch.ipynb +++ b/tts_pytorch.ipynb @@ -63,7 +63,7 @@ "voice_path = tts_model.get_voice_path(voice)\n", "# CFG coef goes here because the model was trained with CFG distillation,\n", "# so it's not _actually_ doing CFG at inference time.\n", - "# Also, if you are generating a dialog, you should have at least two voices in the list.\n", + "# Also, if you are generating a dialog, you should have two voices in the list.\n", "condition_attributes = tts_model.make_condition_attributes(\n", " [voice_path], cfg_coef=2.0\n", ")" @@ -79,8 +79,6 @@ "print(\"Generating audio...\")\n", "\n", "pcms = []\n", - "\n", - "@torch.no_grad()\n", "def _on_frame(frame):\n", " print(\"Step\", len(pcms), end=\"\\r\")\n", " if (frame != -1).all():\n",