Skip to content

Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.

Dev gzf (#1469)

zhifu gaocc59310dbf51
diff --git a/funasr/auto/auto_model.py b/funasr/auto/auto_model.py
index c8cd30c0c..edcede5bf 100644
--- a/funasr/auto/auto_model.py
+++ b/funasr/auto/auto_model.py
@@ -157,7 +157,6 @@ class AutoModel:
kwargs["device"] = device
torch.set_num_threads(kwargs.get("ncpu", 4))
-
# build tokenizer
tokenizer = kwargs.get("tokenizer", None)
diff --git a/funasr/models/bicif_paraformer/model.py b/funasr/models/bicif_paraformer/model.py
index 4802da07d..b93f93a88 100644
--- a/funasr/models/bicif_paraformer/model.py
+++ b/funasr/models/bicif_paraformer/model.py
@@ -367,7 +367,7 @@ class BiCifParaformer(Paraformer):
else:
self.make_pad_mask = sequence_mask(max_seq_len, flip=False)
- self.forward = self._export_forward
+ self.forward = self.export_forward
return self
diff --git a/funasr/models/ct_transformer/model.py b/funasr/models/ct_transformer/model.py
index 88ee86734..9f680fdc8 100644
--- a/funasr/models/ct_transformer/model.py
+++ b/funasr/models/ct_transformer/model.py
@@ -373,7 +373,7 @@ class CTTransformer(torch.nn.Module):
encoder_class = tables.encoder_classes.get(kwargs["encoder"]+"Export")
self.encoder = encoder_class(self.encoder, onnx=is_onnx)
- self.forward = self._export_forward
+ self.forward = self.export_forward
return self
diff --git a/funasr/models/ct_transformer_streaming/model.py b/funasr/models/ct_transformer_streaming/model.py
index 4752c4b5a..129cc95cc 100644
--- a/funasr/models/ct_transformer_streaming/model.py
+++ b/funasr/models/ct_transformer_streaming/model.py
@@ -182,7 +182,7 @@ class CTTransformerStreaming(CTTransformer):
encoder_class = tables.encoder_classes.get(kwargs["encoder"] + "Export")
self.encoder = encoder_class(self.encoder, onnx=is_onnx)
- self.forward = self._export_forward
+ self.forward = self.export_forward
return self
diff --git a/funasr/models/fsmn_vad_streaming/model.py b/funasr/models/fsmn_vad_streaming/model.py
index d06db2093..602cf23d3 100644
--- a/funasr/models/fsmn_vad_streaming/model.py
+++ b/funasr/models/fsmn_vad_streaming/model.py
@@ -647,7 +647,7 @@ class FsmnVADStreaming(nn.Module):
is_onnx = kwargs.get("type", "onnx") == "onnx"
encoder_class = tables.encoder_classes.get(kwargs["encoder"] + "Export")
self.encoder = encoder_class(self.encoder, onnx=is_onnx)
- self.forward = self._export_forward
+ self.forward = self.export_forward
return self
diff --git a/funasr/models/paraformer/model.py b/funasr/models/paraformer/model.py
index f5f0e4e22..2e2a36e06 100644
--- a/funasr/models/paraformer/model.py
+++ b/funasr/models/paraformer/model.py
@@ -575,7 +575,7 @@ class Paraformer(torch.nn.Module):
else:
self.make_pad_mask = sequence_mask(max_seq_len, flip=False)
- self.forward = self._export_forward
+ self.forward = self.export_forward
return self
diff --git a/funasr/models/paraformer_streaming/model.py b/funasr/models/paraformer_streaming/model.py
index 63dba5d88..518fe9369 100644
--- a/funasr/models/paraformer_streaming/model.py
+++ b/funasr/models/paraformer_streaming/model.py
@@ -587,7 +587,6 @@ class ParaformerStreaming(Paraformer):
else:
self.make_pad_mask = sequence_mask(max_seq_len, flip=False)
- self.forward = self._export_forward
import copy
import types
@@ -595,7 +594,7 @@ class ParaformerStreaming(Paraformer):
decoder_model = copy.copy(self)
# encoder
- encoder_model.forward = types.MethodType(ParaformerStreaming._export_encoder_forward, encoder_model)
+ encoder_model.forward = types.MethodType(ParaformerStreaming.export_encoder_forward, encoder_model)
encoder_model.export_dummy_inputs = types.MethodType(ParaformerStreaming.export_encoder_dummy_inputs, encoder_model)
encoder_model.export_input_names = types.MethodType(ParaformerStreaming.export_encoder_input_names, encoder_model)
encoder_model.export_output_names = types.MethodType(ParaformerStreaming.export_encoder_output_names, encoder_model)
@@ -603,7 +602,7 @@ class ParaformerStreaming(Paraformer):
encoder_model.export_name = types.MethodType(ParaformerStreaming.export_encoder_name, encoder_model)
# decoder
- decoder_model.forward = types.MethodType(ParaformerStreaming._export_decoder_forward, decoder_model)
+ decoder_model.forward = types.MethodType(ParaformerStreaming.export_decoder_forward, decoder_model)
decoder_model.export_dummy_inputs = types.MethodType(ParaformerStreaming.export_decoder_dummy_inputs, decoder_model)
decoder_model.export_input_names = types.MethodType(ParaformerStreaming.export_decoder_input_names, decoder_model)
decoder_model.export_output_names = types.MethodType(ParaformerStreaming.export_decoder_output_names, decoder_model)