RuntimeError: User specified an unsupported autocast device_type 'mps'

Question

I am trying to run the distil-whisper training code on my M1 Mac and I face this issue while running the code. It is executed using huggingface accelerate library and I am running into this error every time

RuntimeError: User specified an unsupported autocast device_type 'mps'

and the error comes from file python3.10/site-packages/torch/amp/autocast_mode.py, here i don't see any 'mps' device, but why is that?

This is the config of accelerate in /huggingface/accelerate/default_config.yaml

compute_environment: LOCAL_MACHINE
debug: false
distributed_type: 'NO'
downcast_bf16: 'no'
machine_rank: 0
main_training_function: main
mixed_precision: 'no'
num_machines: 1
num_processes: 1
rdzv_backend: static
same_network: true
tpu_env: []
tpu_use_cluster: false
tpu_use_sudo: false
use_cpu: false