RuntimeError: [enforce fail at CPUAllocator.cpp:65]

[raja1196]

Followed the installation steps, built the Dockerfile (Which had its own hiccups, the fairseq repository does not have base_architecture definition in their models file, will raise a PR for it separately) and ran the code.

Running command
python examples/wav2vec/recognize.py --wav_path /app/data/test.WAV --w2v_path /app/data/wav2vec_small_10m.pt --target_dict_path /app/data/dict.ltr.txt

Error:
RuntimeError: [enforce fail at CPUAllocator.cpp:65] . DefaultCPUAllocator: can't allocate memory: you tried to allocate 314663671488 bytes. Error code 12 (Cannot allocate memory)

Screenshot of the error:

I tried with different models of wave2vec with similar error. Do let me know if more information is needed. Running this on Azure DS VM.

[loretoparisi]

PR for base_architecture definition merged thanks.

[raja1196]

I did a couple of things.

• Changed the memory limit for the docker to 8GB
• used a different audio file

I got the result:
WE THE PEOPLE OF THE UNITED STATES IN ORDER TO FORM A MORE PERFECT UNION ESTABLISH JUSTICE INSURE DOMESTIC TRANQUILLITY PROVIDE FOR THE COMMON DEFENCE

I suspect that the length of the audio file could be a reason it is throwing the error (do let me know if it it is a more prevalent problem). The audio file i was trying to transcribe was around 1hr, compared to the 10 sec clip linked above.

[raja1196]

facebookresearch/fairseq#2651 (comment)

I followed this comment for reference.

[loretoparisi]

@raja1196 thanks we will check this case.

[iman3sh]

I'm trying to run XLRS-53 after fine-tuning with this code but the following error happens. can anyone help me?

/usr/local/lib/python3.8/site-packages/torch/cuda/init.py:52: UserWarning: CUDA initialization: Found no NVIDIA driver on your system. Please check that you have an NVIDIA GPU and installed a driver from http://www.nvidia.com/Download/index.aspx (Triggered internally at /pytorch/c10/cuda/CUDAFunctions.cpp:100.)
return torch._C._cuda_getDeviceCount() > 0
Traceback (most recent call last):
File "src/recognize.py", line 170, in
model = Wav2VecPredictor(args.w2v_path, args.target_dict_path)
File "src/recognize.py", line 122, in init
self._model = self._load_model(w2v_path, self._target_dict)
File "src/recognize.py", line 145, in _load_model
args = base_architecture(w2v["args"])
File "src/recognize.py", line 24, in base_architecture
args.no_pretrained_weights = getattr(args, "no_pretrained_weights", False)