diff --git a/funasr/models/data2vec.py b/funasr/models/data2vec.py
index 2d4711f89..380c137b6 100644
--- a/funasr/models/data2vec.py
+++ b/funasr/models/data2vec.py
@@ -12,7 +12,6 @@ from typing import Tuple
 import torch
 from typeguard import check_argument_types
 
-from funasr.layers.abs_normalize import AbsNormalize
 from funasr.models.preencoder.abs_preencoder import AbsPreEncoder
 from funasr.torch_utils.device_funcs import force_gatherable
 from funasr.models.base_model import FunASRModel
@@ -33,7 +32,7 @@ class Data2VecPretrainModel(FunASRModel):
             self,
             frontend: Optional[torch.nn.Module],
             specaug: Optional[torch.nn.Module],
-            normalize: Optional[AbsNormalize],
+            normalize: Optional[torch.nn.Module],
             preencoder: Optional[AbsPreEncoder],
             encoder: torch.nn.Module,
     ):
diff --git a/funasr/models/e2e_asr.py b/funasr/models/e2e_asr.py
index 1d2b20d8d..950d69960 100644
--- a/funasr/models/e2e_asr.py
+++ b/funasr/models/e2e_asr.py
@@ -13,17 +13,13 @@ from typing import Union
 import torch
 from typeguard import check_argument_types
 
-from funasr.layers.abs_normalize import AbsNormalize
 from funasr.losses.label_smoothing_loss import (
     LabelSmoothingLoss,  # noqa: H301
 )
 from funasr.models.ctc import CTC
 from funasr.models.decoder.abs_decoder import AbsDecoder
-from funasr.models.encoder.abs_encoder import AbsEncoder
-from funasr.models.frontend.abs_frontend import AbsFrontend
 from funasr.models.postencoder.abs_postencoder import AbsPostEncoder
 from funasr.models.preencoder.abs_preencoder import AbsPreEncoder
-from funasr.models.specaug.abs_specaug import AbsSpecAug
 from funasr.models.base_model import FunASRModel
 from funasr.modules.add_sos_eos import add_sos_eos
 from funasr.modules.e2e_asr_common import ErrorCalculator
@@ -46,11 +42,11 @@ class ESPnetASRModel(FunASRModel):
             self,
             vocab_size: int,
             token_list: Union[Tuple[str, ...], List[str]],
-            frontend: Optional[AbsFrontend],
-            specaug: Optional[AbsSpecAug],
-            normalize: Optional[AbsNormalize],
+            frontend: Optional[torch.nn.Module],
+            specaug: Optional[torch.nn.Module],
+            normalize: Optional[torch.nn.Module],
             preencoder: Optional[AbsPreEncoder],
-            encoder: AbsEncoder,
+            encoder: torch.nn.Module,
             postencoder: Optional[AbsPostEncoder],
             decoder: AbsDecoder,
             ctc: CTC,
diff --git a/funasr/models/e2e_asr_mfcca.py b/funasr/models/e2e_asr_mfcca.py
index d04255251..efdd90dc7 100644
--- a/funasr/models/e2e_asr_mfcca.py
+++ b/funasr/models/e2e_asr_mfcca.py
@@ -17,12 +17,8 @@ from funasr.losses.label_smoothing_loss import (
 )
 from funasr.models.ctc import CTC
 from funasr.models.decoder.abs_decoder import AbsDecoder
-from funasr.models.encoder.abs_encoder import AbsEncoder
-from funasr.models.frontend.abs_frontend import AbsFrontend
 from funasr.models.preencoder.abs_preencoder import AbsPreEncoder
-from funasr.models.specaug.abs_specaug import AbsSpecAug
 from funasr.models.base_model import FunASRModel
-from funasr.layers.abs_normalize import AbsNormalize
 from funasr.torch_utils.device_funcs import force_gatherable
 
 
@@ -43,11 +39,11 @@ class MFCCA(FunASRModel):
         self,
         vocab_size: int,
         token_list: Union[Tuple[str, ...], List[str]],
-        frontend: Optional[AbsFrontend],
-        specaug: Optional[AbsSpecAug],
-        normalize: Optional[AbsNormalize],
+        frontend: Optional[torch.nn.Module],
+        specaug: Optional[torch.nn.Module],
+        normalize: Optional[torch.nn.Module],
         preencoder: Optional[AbsPreEncoder],
-        encoder: AbsEncoder,
+        encoder: torch.nn.Module,
         decoder: AbsDecoder,
         ctc: CTC,
         rnnt_decoder: None,
diff --git a/funasr/models/e2e_asr_paraformer.py b/funasr/models/e2e_asr_paraformer.py
index cf5c16d15..288f46995 100644
--- a/funasr/models/e2e_asr_paraformer.py
+++ b/funasr/models/e2e_asr_paraformer.py
@@ -12,19 +12,15 @@ import random
 import numpy as np
 from typeguard import check_argument_types
 
-from funasr.layers.abs_normalize import AbsNormalize
 from funasr.losses.label_smoothing_loss import (
     LabelSmoothingLoss,  # noqa: H301
 )
 from funasr.models.ctc import CTC
 from funasr.models.decoder.abs_decoder import AbsDecoder
 from funasr.models.e2e_asr_common import ErrorCalculator
-from funasr.models.encoder.abs_encoder import AbsEncoder
-from funasr.models.frontend.abs_frontend import AbsFrontend
 from funasr.models.postencoder.abs_postencoder import AbsPostEncoder
 from funasr.models.predictor.cif import mae_loss
 from funasr.models.preencoder.abs_preencoder import AbsPreEncoder
-from funasr.models.specaug.abs_specaug import AbsSpecAug
 from funasr.models.base_model import FunASRModel
 from funasr.modules.add_sos_eos import add_sos_eos
 from funasr.modules.nets_utils import make_pad_mask, pad_list
@@ -53,11 +49,11 @@ class Paraformer(FunASRModel):
             self,
             vocab_size: int,
             token_list: Union[Tuple[str, ...], List[str]],
-            frontend: Optional[AbsFrontend],
-            specaug: Optional[AbsSpecAug],
-            normalize: Optional[AbsNormalize],
+            frontend: Optional[torch.nn.Module],
+            specaug: Optional[torch.nn.Module],
+            normalize: Optional[torch.nn.Module],
             preencoder: Optional[AbsPreEncoder],
-            encoder: AbsEncoder,
+            encoder: torch.nn.Module,
             postencoder: Optional[AbsPostEncoder],
             decoder: AbsDecoder,
             ctc: CTC,
@@ -620,11 +616,11 @@ class ParaformerBert(Paraformer):
             self,
             vocab_size: int,
             token_list: Union[Tuple[str, ...], List[str]],
-            frontend: Optional[AbsFrontend],
-            specaug: Optional[AbsSpecAug],
-            normalize: Optional[AbsNormalize],
+            frontend: Optional[torch.nn.Module],
+            specaug: Optional[torch.nn.Module],
+            normalize: Optional[torch.nn.Module],
             preencoder: Optional[AbsPreEncoder],
-            encoder: AbsEncoder,
+            encoder: torch.nn.Module,
             postencoder: Optional[AbsPostEncoder],
             decoder: AbsDecoder,
             ctc: CTC,
@@ -898,11 +894,11 @@ class BiCifParaformer(Paraformer):
         self,
         vocab_size: int,
         token_list: Union[Tuple[str, ...], List[str]],
-        frontend: Optional[AbsFrontend],
-        specaug: Optional[AbsSpecAug],
-        normalize: Optional[AbsNormalize],
+        frontend: Optional[torch.nn.Module],
+        specaug: Optional[torch.nn.Module],
+        normalize: Optional[torch.nn.Module],
         preencoder: Optional[AbsPreEncoder],
-        encoder: AbsEncoder,
+        encoder: torch.nn.Module,
         postencoder: Optional[AbsPostEncoder],
         decoder: AbsDecoder,
         ctc: CTC,
@@ -1111,11 +1107,11 @@ class ContextualParaformer(Paraformer):
             self,
             vocab_size: int,
             token_list: Union[Tuple[str, ...], List[str]],
-            frontend: Optional[AbsFrontend],
-            specaug: Optional[AbsSpecAug],
-            normalize: Optional[AbsNormalize],
+            frontend: Optional[torch.nn.Module],
+            specaug: Optional[torch.nn.Module],
+            normalize: Optional[torch.nn.Module],
             preencoder: Optional[AbsPreEncoder],
-            encoder: AbsEncoder,
+            encoder: torch.nn.Module,
             postencoder: Optional[AbsPostEncoder],
             decoder: AbsDecoder,
             ctc: CTC,