fastt5 not working with FastAPI gunicorn and docker

[kklivil]

I am using fastt5 with FastAPI web framework and gunicorn as a server within a docker container. The server doesn't startup completely i.e. it hangs during the startup process.
command to start the server: gunicorn app.my_app:app --bind 0.0.0.0:${PORT} --reload --timeout 120 --access-logfile -

requirements.txt:

anyio==3.6.2
certifi==2022.9.24
charset-normalizer==2.1.1
click==8.1.3
fastapi==0.85.1
filelock==3.8.0
h11==0.14.0
huggingface-hub==0.10.1
idna==3.4
numpy==1.23.4
omegaconf==2.2.3
packaging==21.3
pydantic==1.10.2
pyparsing==3.0.9
PyYAML==6.0
regex==2022.9.13
requests==2.28.1
sentencepiece==0.1.97
sniffio==1.3.0
starlette==0.20.4
tokenizers==0.13.1
torch==1.12.1
tqdm==4.64.1
transformers==4.23.1
typing_extensions==4.4.0
urllib3==1.26.12
uvicorn==0.19.0
gunicorn==20.1.0
httptools==0.5.0
python-dotenv==0.21.0
uvloop==0.17.0
watchfiles==0.18.0
websockets==10.4
fastt5==0.1.4
six==1.16.0

There is NO error in the output during the start-up. It just hangs.

[2022-11-17 14:19:38 +0000] [7] [INFO] Listening at: http://0.0.0.0:8000 (7)
[2022-11-17 14:19:38 +0000] [7] [INFO] Using worker: sync
[2022-11-17 14:19:38 +0000] [8] [INFO] Booting worker with pid: 8
Downloading: 100%|██████████| 1.20k/1.20k [00:00<00:00, 473kB/s]
Downloading: 100%|██████████| 242M/242M [00:28<00:00, 8.41MB/s] 
In-place op on output of tensor.shape. See https://pytorch.org/docs/master/onnx.html#avoid-inplace-operations-when-using-tensor-shape-in-tracing-mode
In-place op on output of tensor.shape. See https://pytorch.org/docs/master/onnx.html#avoid-inplace-operations-when-using-tensor-shape-in-tracing-mode
Exporting to onnx... |################################| 3/3
Quantizing... |################################| 3/3
Downloading:   0%|          | 0.00/792k [00:00<?, ?B/s]Setting up onnx model...
Done!
Downloading: 100%|██████████| 792k/792k [00:00<00:00, 1.41MB/s] 

Could you please have a look into this?

[cm2435]

Bumping this. same issue