audio

audio copied to clipboard

version bump

1

PLEASE NOTE THAT THE TORCHAUDIO REPOSITORY IS NO LONGER ACTIVELY MONITORED. You may not get a response. For open discussions, visit https://discuss.pytorch.org/.

PaliC

torchaudio.load not loading all the frames

1

### 🐛 Describe the bug torchaudio.load not loading all the frames in the latest version(2.2.1). Example audio can be downloaded from [here](https://freesound.org/people/sumsa/sounds/46673/download/46673__sumsa__harddisk-read.wav) ``` import torchaudio file = "harddisk_operation.wav" audio, sr...

ashinkajay

Add HIP path for ROCM 6.0

2

Adding HIP-cmake path for ROCm 6.0 to CMAKE_MODULE_PATH to find HIP module.

ramcherukuri

Get cudaErrorIllegalAddress when running ctc decoder on H100

2

### 🐛 Describe the bug I am getting this error when running the sample code form pytorch [https://pytorch.org/audio/2.2.0/tutorials/asr_inference_with_cuda_ctc_decoder_tutorial.html?highlight=ctc](url) But everything is okay when running on both 4090/3090 machine. What kind...

Ray-Leung

torchaudio load opus failed

1

### 🐛 Describe the bug ```python >>> torchaudio.backend.soundfile_backend.load('test.opus') (tensor([[-5.2875e-12, 2.0960e-11, 2.4085e-11, ..., 2.9903e-06, -6.7451e-06, 3.6049e-05]]), 16000) >>> torchaudio.backend.sox_io_backend.load('test.opus') Traceback (most recent call last): File "", line 1, in File "/usr/local/lib/miniconda3/envs/cloud-ai-lab/lib/python3.8/site-packages/torchaudio/backend/_sox_io_backend.py",...

Mddct

Reduce memory usage of forced alignment on CPU

5

In the forced alignment c++ code, `backPtr` is an `int8` tensor while only storing the values 0,1, and 2 which can be effectively stored using only 2 bits instead of...

MahmoudAshraf97

Update asr_inference_with_ctc_decoder_tutorial.py

3

fixes typo in lexicon filename PLEASE NOTE THAT THE TORCHAUDIO REPOSITORY IS NO LONGER ACTIVELY MONITORED. You may not get a response. For open discussions, visit https://discuss.pytorch.org/.

kappalabs

MAC M3 audio backend no longer appearing

4

### 🐛 Description I get "RuntimeError: Couldn't find appropriate backend to handle uri and format None." when processing basic audio data on my mac but not on my colab notebook....

tval2

### 🐛 Describe the bug s = StreamWriter(dst="-", format="audiotoolbox") s.add_audio_stream(sample_rate, num_channels, format="s16") --------------------------------------------------------------------------- RuntimeError Traceback (most recent call last) Cell In[5], line 2 1 # Configure StreamWriter to write to...

liuliujiang

frame offset + num frames to utilize http range header

1

### 🚀 The feature If load is called with a url e.g. ```py torchaudio.load(url, frame_offset=16000, num_frames=16000) ``` And the url supports range headers, is it not possible to only get...

mogwai