Segmentation fault when running python transformer_infer.py

[CircleCly]

Hi, I am on an AWS EC2 Machine (Amazon Linux 2), and I followed the instructions in the setup and are able to install the library dependencies. I also downloaded all the checkpoints on my machine. However, when I run python transformer_infer.py, I get the following message and a segmentation fault:

/opt/conda/envs/pheme3/lib/python3.10/site-packages/pyannote/audio/core/io.py:43: UserWarning: torchaudio._backend.set_audio_backend has been deprecated. With dispatcher enabled, this function is no-op. You can remove the function call.
  torchaudio.set_audio_backend("soundfile")
/opt/conda/envs/pheme3/lib/python3.10/site-packages/torch_audiomentations/utils/io.py:27: UserWarning: torchaudio._backend.set_audio_backend has been deprecated. With dispatcher enabled, this function is no-op. You can remove the function call.
  torchaudio.set_audio_backend("soundfile")
DEBUG:fsspec.local:open file: /home/ec2-user/wam/pheme/ckpt/s2a/s2a.ckpt
Non-A100 GPU detected, using math or mem efficient attention if input tensor is on cuda
DEBUG:urllib3.connectionpool:Starting new HTTPS connection (1): huggingface.co:443
DEBUG:urllib3.connectionpool:https://huggingface.co:443 "HEAD /pyannote/embedding/resolve/main/pytorch_model.bin HTTP/1.1" 302 0
DEBUG:urllib3.connectionpool:https://huggingface.co:443 "HEAD /pyannote/embedding/resolve/main/config.yaml HTTP/1.1" 200 0
DEBUG:fsspec.local:open file: /home/ec2-user/.cache/torch/pyannote/models--pyannote--embedding/snapshots/c6335d8f1cd77b30084387468a6cf26fea90009b/pytorch_model.bin
DEBUG:fsspec.local:open file: /home/ec2-user/.cache/torch/pyannote/models--pyannote--embedding/snapshots/c6335d8f1cd77b30084387468a6cf26fea90009b/pytorch_model.bin
Lightning automatically upgraded your loaded checkpoint from v1.2.7 to v2.1.2. To apply the upgrade to your files permanently, run `python -m pytorch_lightning.utilities.upgrade_checkpoint ../../.cache/torch/pyannote/models--pyannote--embedding/snapshots/c6335d8f1cd77b30084387468a6cf26fea90009b/pytorch_model.bin`
Model was trained with pyannote.audio 0.0.1, yours is 3.1.1. Bad things might happen unless you revert pyannote.audio to 0.x.
Model was trained with torch 1.8.1+cu102, yours is 2.1.1+cu121. Bad things might happen unless you revert torch to 1.x.
Segmentation fault

I have tried debugging the python script and stepping into function calls.

This is the last non-library function call I was at before erroring, which was in the TextTokenizer class in semantic_dataset.py. From this I believe the phonemizer backend is erroring.

I tried stepping into the phonemizer library as well, and the program errors out after this line is executed:

And this is my call stack:

Any help on the issue would be greatly appreciated!