From 7ea3836893bfdf1aac03952bb1ff2da2c6ef6e57 Mon Sep 17 00:00:00 2001
From: 嘉渊 <wangjiaming.wjm@alibaba-inc.com>
Date: 星期二, 01 八月 2023 14:18:32 +0800
Subject: [PATCH] update

---
 funasr/bin/asr_infer.py |   52 +---------------------------------------------------
 1 files changed, 1 insertions(+), 51 deletions(-)

diff --git a/funasr/bin/asr_infer.py b/funasr/bin/asr_infer.py
index 0ce8dd8..02ca63d 100644
--- a/funasr/bin/asr_infer.py
+++ b/funasr/bin/asr_infer.py
@@ -22,9 +22,7 @@
 import requests
 import torch
 from packaging.version import parse as V
-from typeguard import check_argument_types
-from typeguard import check_return_type
-from  funasr.build_utils.build_model_from_file import build_model_from_file
+from funasr.build_utils.build_model_from_file import build_model_from_file
 from funasr.models.e2e_asr_contextual_paraformer import NeatContextualParaformer
 from funasr.models.e2e_asr_paraformer import BiCifParaformer, ContextualParaformer
 from funasr.models.frontend.wav_frontend import WavFrontend, WavFrontendOnline
@@ -78,7 +76,6 @@
             frontend_conf: dict = None,
             **kwargs,
     ):
-        assert check_argument_types()
 
         # 1. Build ASR model
         scorers = {}
@@ -192,7 +189,6 @@
             text, token, token_int, hyp
 
         """
-        assert check_argument_types()
 
         # Input as audio signal
         if isinstance(speech, np.ndarray):
@@ -248,7 +244,6 @@
                 text = None
             results.append((text, token, token_int, hyp))
 
-        assert check_return_type(results)
         return results
 
 
@@ -289,7 +284,6 @@
             decoding_ind: int = 0,
             **kwargs,
     ):
-        assert check_argument_types()
 
         # 1. Build ASR model
         scorers = {}
@@ -415,7 +409,6 @@
                 text, token, token_int, hyp
 
         """
-        assert check_argument_types()
 
         # Input as audio signal
         if isinstance(speech, np.ndarray):
@@ -522,7 +515,6 @@
                                                                vad_offset=begin_time)
                 results.append((text, token, token_int, hyp, timestamp, enc_len_batch_total, lfr_factor))
 
-        # assert check_return_type(results)
         return results
 
     def generate_hotwords_list(self, hotword_list_or_file):
@@ -662,7 +654,6 @@
             hotword_list_or_file: str = None,
             **kwargs,
     ):
-        assert check_argument_types()
 
         # 1. Build ASR model
         scorers = {}
@@ -782,7 +773,6 @@
                 text, token, token_int, hyp
 
         """
-        assert check_argument_types()
         results = []
         cache_en = cache["encoder"]
         if speech.shape[1] < 16 * 60 and cache_en["is_final"]:
@@ -877,7 +867,6 @@
 
                 results.append(postprocessed_result)
 
-        # assert check_return_type(results)
         return results
 
 
@@ -918,7 +907,6 @@
             frontend_conf: dict = None,
             **kwargs,
     ):
-        assert check_argument_types()
 
         # 1. Build ASR model
         scorers = {}
@@ -1042,7 +1030,6 @@
             text, token, token_int, hyp
 
         """
-        assert check_argument_types()
 
         # Input as audio signal
         if isinstance(speech, np.ndarray):
@@ -1110,7 +1097,6 @@
                 text = None
             results.append((text, token, token_int, hyp))
 
-        assert check_return_type(results)
         return results
 
 
@@ -1149,7 +1135,6 @@
             streaming: bool = False,
             **kwargs,
     ):
-        assert check_argument_types()
 
         # 1. Build ASR model
         scorers = {}
@@ -1254,7 +1239,6 @@
             text, token, token_int, hyp
 
         """
-        assert check_argument_types()
         # Input as audio signal
         if isinstance(speech, np.ndarray):
             speech = torch.tensor(speech)
@@ -1304,7 +1288,6 @@
                 text = None
             results.append((text, token, token_int, hyp))
 
-        assert check_return_type(results)
         return results
 
 
@@ -1361,7 +1344,6 @@
         """Construct a Speech2Text object."""
         super().__init__()
 
-        assert check_argument_types()
         asr_model, asr_train_args = build_model_from_file(
             asr_train_config, asr_model_file, cmvn_file, device
         )
@@ -1540,7 +1522,6 @@
         Returns:
             nbest_hypothesis: N-best hypothesis.
         """
-        assert check_argument_types()
 
         if isinstance(speech, np.ndarray):
             speech = torch.tensor(speech)
@@ -1572,7 +1553,6 @@
         Returns:
             nbest_hypothesis: N-best hypothesis.
         """
-        assert check_argument_types()
 
         if isinstance(speech, np.ndarray):
             speech = torch.tensor(speech)
@@ -1614,35 +1594,8 @@
                 text = None
             results.append((text, token, token_int, hyp))
 
-            assert check_return_type(results)
 
         return results
-
-    @staticmethod
-    def from_pretrained(
-            model_tag: Optional[str] = None,
-            **kwargs: Optional[Any],
-    ) -> Speech2Text:
-        """Build Speech2Text instance from the pretrained model.
-        Args:
-            model_tag: Model tag of the pretrained models.
-        Return:
-            : Speech2Text instance.
-        """
-        if model_tag is not None:
-            try:
-                from espnet_model_zoo.downloader import ModelDownloader
-
-            except ImportError:
-                logging.error(
-                    "`espnet_model_zoo` is not installed. "
-                    "Please install via `pip install -U espnet_model_zoo`."
-                )
-                raise
-            d = ModelDownloader()
-            kwargs.update(**d.download_and_unpack(model_tag))
-
-        return Speech2TextTransducer(**kwargs)
 
 
 class Speech2TextSAASR:
@@ -1681,7 +1634,6 @@
             frontend_conf: dict = None,
             **kwargs,
     ):
-        assert check_argument_types()
 
         # 1. Build ASR model
         scorers = {}
@@ -1799,7 +1751,6 @@
             text, text_id, token, token_int, hyp
 
         """
-        assert check_argument_types()
 
         # Input as audio signal
         if isinstance(speech, np.ndarray):
@@ -1892,5 +1843,4 @@
 
             results.append((text, text_id, token, token_int, hyp))
 
-        assert check_return_type(results)
         return results

--
Gitblit v1.9.1