From 9c884c566ff1a7c26f0f28e8d8ad4deb281a954d Mon Sep 17 00:00:00 2001
From: 语帆 <yf352572@alibaba-inc.com>
Date: 星期一, 04 三月 2024 16:48:01 +0800
Subject: [PATCH] atsr

---
 examples/industrial_data_pretraining/lcbnet/README.md |   77 +++++++++++++++++++++++++++++++++++++-
 1 files changed, 75 insertions(+), 2 deletions(-)

diff --git a/examples/industrial_data_pretraining/lcbnet/README.md b/examples/industrial_data_pretraining/lcbnet/README.md
index 4273ec0..9235886 100644
--- a/examples/industrial_data_pretraining/lcbnet/README.md
+++ b/examples/industrial_data_pretraining/lcbnet/README.md
@@ -75,9 +75,9 @@
 闅忕潃鍦ㄧ嚎浼氳鍜岃绋嬭秺鏉ヨ秺鏅亶锛屽浣曞埄鐢ㄨ棰戝够鐏墖涓赴瀵岀殑鏂囨湰淇℃伅鏉ユ敼鍠勮闊宠瘑鍒紙Automatic聽 Speech Recognition锛� ASR锛夐潰涓寸潃鏂扮殑鎸戞垬銆傝棰戜腑鐨勫够鐏墖涓庤闊冲疄鏃跺悓姝ワ紝鐩告瘮浜庣粺涓�鐨勭█鏈夎瘝鍒楄〃锛岃兘澶熸彁渚涙洿闀跨殑涓婁笅鏂囩浉鍏充俊鎭�傚洜姝わ紝鎴戜滑鎻愬嚭浜嗕竴绉嶅垱鏂扮殑闀夸笂涓嬫枃鍋忕疆缃戠粶锛圠CB-net锛夛紝鐢ㄤ簬闊抽-瑙嗚璇煶璇嗗埆锛圓udio-Visual Speech Recognition锛孉VSR锛夛紝浠ユ洿濂藉湴鍒╃敤瑙嗛涓殑闀挎椂涓婁笅鏂囦俊鎭��
 
 <p align="center">
-<img src="fig/lcbnet1.png" alt="AVSR鏁翠綋娴佺▼妗嗘灦"  width="500" />
+<img src="fig/lcbnet1.png" alt="AVSR鏁翠綋娴佺▼妗嗘灦"  width="800" />
 <p align="center">
-<img src="fig/lcbnet2.png" alt="LCB-NET妯″瀷缁撴瀯"  width="500" />
+<img src="fig/lcbnet2.png" alt="LCB-NET妯″瀷缁撴瀯"  width="800" />
 
 
 鍏蜂綋鏉ヨ锛屾垜浠鍏堜娇鐢∣CR鎶�鏈潵妫�娴嬪拰璇嗗埆骞荤伅鐗囦腑鐨勬枃鏈唴瀹癸紝鍏舵鎴戜滑閲囩敤鍏抽敭璇嶆彁鍙栨妧鏈潵鑾峰彇鏂囨湰鍐呭涓殑鍏抽敭璇嶇煭璇�傛渶鍚庯紝鎴戜滑灏嗗叧閿瘝鎷兼帴鎴愰暱涓婁笅鏂囨枃鏈拰闊抽鍚屾椂杈撳叆鍒版垜浠殑LCB-net妯″瀷涓繘琛岃瘑鍒�傝�孡CB-net妯″瀷閲囩敤浜嗗弻缂栫爜鍣ㄧ粨鏋勶紝鍚屾椂寤烘ā闊抽鍜岄暱涓婁笅鏂囨枃鏈俊鎭�傛澶栵紝鎴戜滑杩樺紩鍏ヤ簡涓�涓樉寮忕殑鍋忕疆璇嶉娴嬫ā鍧楋紝閫氳繃浣跨敤浜屽厓浜ゅ弶鐔碉紙BCE锛夋崯澶卞嚱鏁版樉寮忛娴嬮暱涓婁笅鏂囨枃鏈腑鍦ㄩ煶棰戜腑鍑虹幇鐨勫叧閿亸缃瘝銆傛澶栵紝涓哄寮篖CB-net鐨勬硾鍖栬兘鍔涘拰绋冲仴鎬э紝鎴戜滑杩橀噰鐢ㄤ簡鍔ㄦ�佺殑鍏抽敭璇嶆ā鎷熺瓥鐣ャ�傚疄楠岃瘉鏄庯紝鎴戜滑鎻愬嚭鐨凩CB-net鐑瘝妯″瀷锛屼笉浠呰兘澶熸彁鍗囧叧閿瘝鐨勮瘑鍒晥鏋滐紝鍚屾椂涔熻兘澶熸彁鍗囬潪鍏抽敭璇嶇殑璇嗗埆鏁堟灉銆傚叿浣撳疄楠岀粨鏋滃涓嬫墍绀猴細
@@ -91,7 +91,80 @@
 
 
 
+## 鍩轰簬ModelScope杩涜鎺ㄧ悊
 
+- 鎺ㄧ悊鏀寔闊抽鏍煎紡濡備笅锛�
+  - wav鏂囦欢璺緞锛屼緥濡傦細data/test/asr_example.wav
+  - pcm鏂囦欢璺緞锛屼緥濡傦細data/test/asr_example.pcm
+  - ark鏂囦欢璺緞锛屼緥濡傦細data/test/data.ark
+  - wav鏂囦欢url锛屼緥濡傦細https://www.modelscope.cn/api/v1/models/iic/LCB-NET/repo?Revision=master&FilePath=example/asr_example.wav
+  - wav浜岃繘鍒舵暟鎹紝鏍煎紡bytes锛屼緥濡傦細鐢ㄦ埛鐩存帴浠庢枃浠堕噷璇诲嚭bytes鏁版嵁鎴栬�呮槸楹﹀厠椋庡綍鍑篵ytes鏁版嵁銆�
+  - 宸茶В鏋愮殑audio闊抽锛屼緥濡傦細audio, rate = soundfile.read("asr_example_zh.wav")锛岀被鍨嬩负numpy.ndarray鎴栬�卼orch.Tensor銆�
+  - wav.scp鏂囦欢锛岄渶绗﹀悎濡備笅瑕佹眰(浠ヤ笅鍒嗗埆涓簊ound鍜宬aldi_ark鏍煎紡)锛�
+
+```sh
+cat wav.scp
+asr_example1  data/test/asr_example1.wav
+asr_example2  data/test/asr_example2.wav
+
+cat wav.scp
+asr_example1  data/test/data_wav.ark:22
+asr_example2  data/test/data_wav.ark:90445
+...
+```
+
+- 鎺ㄧ悊鏀寔OCR棰勬祴鏂囨湰鏍煎紡濡備笅锛�
+  - ocr.txt鏂囦欢锛岄渶绗﹀悎濡備笅瑕佹眰锛�
+```sh
+cat ocr.txt
+asr_example1  ANIMAL <blank> RIGHTS <blank> MANAGER <blank> PLOEG
+asr_example2  UNIVERSITY <blank> CAMPUS <blank> DEANO
+...
+```
+
+- 鑻ヨ緭鍏ユ牸寮弚av鏂囦欢鍜宱cr鏂囦欢鍧囦负url锛宎pi璋冪敤鏂瑰紡鍙弬鑰冨涓嬭寖渚嬶細
+
+```python
+from funasr import AutoModel
+
+model = AutoModel(model="iic/LCB-NET",
+                  model_revision="v2.0.0")
+res = model.generate(input=("https://www.modelscope.cn/api/v1/models/iic/LCB-NET/repo?Revision=master&FilePath=example/asr_example.wav","https://www.modelscope.cn/api/v1/models/iic/LCB-NET/repo?Revision=master&FilePath=example/ocr.txt"),data_type=("sound", "text"))
+```
+
+
+## 澶嶇幇璁烘枃涓殑缁撴灉
+```python
+python -m funasr.bin.inference \
+        --config-path=${file_dir} \
+        --config-name="config.yaml" \
+        ++init_param=${file_dir}/model.pt \
+        ++tokenizer_conf.token_list=${file_dir}/tokens.txt \
+        ++input=[${_logdir}/wav.scp,${_logdir}/ocr.txt] \
+        +data_type='["kaldi_ark", "text"]' \
+        ++tokenizer_conf.bpemodel=${file_dir}/bpe.pt \
+        ++output_dir="${inference_dir}/results" \
+        ++device="${inference_device}" \
+        ++ncpu=1 \
+        ++disable_log=true
+
+```
+
+
+璇嗗埆缁撴灉杈撳嚭璺緞缁撴瀯濡備笅锛�
+
+```sh
+tree output_dir/
+output_dir/
+鈹斺攢鈹� 1best_recog
+    鈹溾攢鈹� text
+    鈹斺攢鈹� token
+```
+
+token锛氳闊宠瘑鍒粨鏋滄枃浠�
+
+鍙互浣跨敤funasr閲岄潰鎻愪緵鐨剅un_bwer_recall.sh璁$畻WER銆丅WER銆乁WER鍜孯ecall銆�
+璇︾粏鑴氭湰鍙互鍙傝�僨unasr閲岄潰鐨刣emo.sh鑴氭湰锛岄渶瑕佹敞鎰忕殑鏄綘闇�瑕佷慨鏀逛竴涓媔ic/LCB-NET/conf.yaml涓瑿MVN(stats_file)鐨勮矾寰勫拰iic/LCB-NET/dev/wav.scp閲岄潰ark鐨勮矾寰勶紝淇敼涓轰綘鑷繁鏈湴鐨勮矾寰勶紝鐒跺悗璺戣В鐮併��
 
 ## 鐩稿叧璁烘枃浠ュ強寮曠敤淇℃伅
 

--
Gitblit v1.9.1