Learning rate does not change for SpaCy NER transformer model

Question

I am training a transformer NER model with GPU, I am changing the learning rate in the config.cfg but it does not apply when I run.

\[training.optimizer.learn_rate\]
@schedules = "warmup_linear.v1"
warmup_steps = 500
total_steps = 20000
initial_rate = 0.00002

but when I run train the model :

Training pipeline
ℹ Pipeline: \['transformer', 'ner'\]
ℹ Initial learn rate: 0.0

I have also tried to give the parameter during the training, but that does not apply too.

python -m spacy train ./config.cfg --output ./model --paths.train ./train.spacy --paths.dev ./test.spacy --gpu-id 0 --training.optimizer.learn_rate.initial_rate 0.00002

config.cfg file

[training.optimizer.learn_rate]
@schedules = "warmup_linear.v1"
warmup_steps = 300
total_steps = 20000
initial_rate = 0.00001