提交

kongdeqiang

2026-03-13 28ccfbfc51068a663a80764e14074df5edf2b5ba

 funasr/models/e_branchformer/encoder.py

@@ -44,6 +44,7 @@
)
from funasr.register import tables


class EBranchformerEncoderLayer(torch.nn.Module):
    """E-Branchformer encoder layer module.

@@ -174,6 +175,7 @@

        return x, mask


@tables.register("encoder_classes", "EBranchformerEncoder")
class EBranchformerEncoder(nn.Module):
    """E-Branchformer encoder module."""
@@ -232,9 +234,7 @@
        elif pos_enc_layer_type == "legacy_rel_pos":
            assert attention_layer_type == "legacy_rel_selfattn"
            pos_enc_class = LegacyRelPositionalEncoding
            logging.warning(
                "Using legacy_rel_pos and it will be deprecated in the future."
            )
            logging.warning("Using legacy_rel_pos and it will be deprecated in the future.")
        else:
            raise ValueError("unknown pos_enc_layer: " + pos_enc_layer_type)

@@ -320,9 +320,7 @@
                output_size,
                attention_dropout_rate,
            )
            logging.warning(
                "Using legacy_rel_selfattn and it will be deprecated in the future."
            )
            logging.warning("Using legacy_rel_selfattn and it will be deprecated in the future.")
        elif attention_layer_type == "rel_selfattn":
            assert pos_enc_layer_type == "rel_pos"
            encoder_selfattn_layer = RelPositionMultiHeadedAttention
@@ -360,9 +358,7 @@
                encoder_selfattn_layer(*encoder_selfattn_layer_args),
                cgmlp_layer(*cgmlp_layer_args),
                positionwise_layer(*positionwise_layer_args) if use_ffn else None,
                positionwise_layer(*positionwise_layer_args)
                if use_ffn and macaron_ffn
                else None,
                positionwise_layer(*positionwise_layer_args) if use_ffn and macaron_ffn else None,
                dropout_rate,
                merge_conv_kernel,
            ),