From 9bf54716138d323f3479efd4daee2355e5d71509 Mon Sep 17 00:00:00 2001
From: zxcd <228587199@qq.com>
Date: Wed, 15 Mar 2023 11:15:02 +0800
Subject: [PATCH] optional tokenizer and fix some doc. (#3042)

---
 examples/librispeech/asr3/local/data.sh        | 0
 examples/librispeech/asr3/local/test.sh        | 0
 examples/librispeech/asr3/local/test_wav.sh    | 0
 examples/librispeech/asr3/local/train.sh       | 0
 examples/librispeech/asr3/run.sh               | 2 +-
 paddlespeech/s2t/exps/wav2vec2/bin/test_wav.py | 5 +++--
 6 files changed, 4 insertions(+), 3 deletions(-)
 mode change 100644 => 100755 examples/librispeech/asr3/local/data.sh
 mode change 100644 => 100755 examples/librispeech/asr3/local/test.sh
 mode change 100644 => 100755 examples/librispeech/asr3/local/test_wav.sh
 mode change 100644 => 100755 examples/librispeech/asr3/local/train.sh

diff --git a/examples/librispeech/asr3/local/data.sh b/examples/librispeech/asr3/local/data.sh
old mode 100644
new mode 100755
diff --git a/examples/librispeech/asr3/local/test.sh b/examples/librispeech/asr3/local/test.sh
old mode 100644
new mode 100755
diff --git a/examples/librispeech/asr3/local/test_wav.sh b/examples/librispeech/asr3/local/test_wav.sh
old mode 100644
new mode 100755
diff --git a/examples/librispeech/asr3/local/train.sh b/examples/librispeech/asr3/local/train.sh
old mode 100644
new mode 100755
diff --git a/examples/librispeech/asr3/run.sh b/examples/librispeech/asr3/run.sh
index 05ad505c..f52266a1 100644
--- a/examples/librispeech/asr3/run.sh
+++ b/examples/librispeech/asr3/run.sh
@@ -6,7 +6,7 @@ set -e
 
 gpus=0
 stage=0
-stop_stage=0
+stop_stage=4
 conf_path=conf/wav2vec2ASR.yaml
 ips=            #xx.xx.xx.xx,xx.xx.xx.xx
 decode_conf_path=conf/tuning/decode.yaml
diff --git a/paddlespeech/s2t/exps/wav2vec2/bin/test_wav.py b/paddlespeech/s2t/exps/wav2vec2/bin/test_wav.py
index 2416db7e..0295713f 100644
--- a/paddlespeech/s2t/exps/wav2vec2/bin/test_wav.py
+++ b/paddlespeech/s2t/exps/wav2vec2/bin/test_wav.py
@@ -34,8 +34,9 @@ class Wav2vec2Infer():
         self.args = args
         self.config = config
         self.audio_file = args.audio_file
+        self.tokenizer = config.get("tokenizer", None)
 
-        if self.config.tokenizer:
+        if self.tokenizer:
             self.text_feature = AutoTokenizer.from_pretrained(
                 self.config.tokenizer)
         else:
@@ -72,7 +73,7 @@ class Wav2vec2Infer():
                 text_feature=self.text_feature,
                 decoding_method=decode_config.decoding_method,
                 beam_size=decode_config.beam_size,
-                tokenizer=self.config.tokenizer, )
+                tokenizer=self.tokenizer, )
             rsl = result_transcripts[0]
             utt = Path(self.audio_file).name
             logger.info(f"hyp: {utt} {rsl}")