Training model Fail "assert aligner.has_params()"

[tinhchuquang]

When I training model with paramater và verification same

    --strategy=ddp \
    --devices=4 \
    --precision="32-true" \
      train.py trainers.prefix=vit_base_KPRPE_WF4M_adaface \
      trainers.num_gpu=4 \
      trainers.batch_size=900 \
      trainers.gradient_acc=12 \
      trainers.num_workers=8 \
      trainers.precision='32-true' \
      trainers.float32_matmul_precision='high' \
      dataset=configs/webface42m_ldmktheta_RA10.yaml \
      data_augs=configs/gridsample_v1.yaml \
      models=vit_kprpe/configs/v1_base_kprpe_splithead_unshared.yaml \
      pipelines=configs/train_keypoint_model_cls.yaml \
      evaluations=configs/quick.yaml \
      classifiers=configs/partial_fc.yaml \
      optims=configs/step_sgd.yaml \
      losses=configs/adaface.yaml \
      trainers.skip_final_eval=False``` 
      -> Error. 
      DDP parameter verification failed ❌
      DDP parameter verification failed ❌
     DDP parameter verification failed ❌
      DDP parameter verification failed ❌
assert aligner.has_params().
Can you help me fix it? thanks.