From c45b89970645ae4c05bd29114b0c1b1e31f70cbc Mon Sep 17 00:00:00 2001
From: haoneng.lhn <haoneng.lhn@alibaba-inc.com>
Date: 星期一, 25 九月 2023 16:55:12 +0800
Subject: [PATCH] update paraformer online README
---
egs_modelscope/asr/TEMPLATE/README.md | 11 +++++++----
egs_modelscope/asr/TEMPLATE/README_zh.md | 11 +++++++----
2 files changed, 14 insertions(+), 8 deletions(-)
diff --git a/egs_modelscope/asr/TEMPLATE/README.md b/egs_modelscope/asr/TEMPLATE/README.md
index a8cb486..ef724b5 100644
--- a/egs_modelscope/asr/TEMPLATE/README.md
+++ b/egs_modelscope/asr/TEMPLATE/README.md
@@ -27,15 +27,18 @@
inference_pipeline = pipeline(
task=Tasks.auto_speech_recognition,
model='damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online',
- model_revision='v1.0.6',
+ model_revision='v1.0.7',
update_model=False,
mode='paraformer_streaming'
)
import soundfile
speech, sample_rate = soundfile.read("example/asr_example.wav")
-chunk_size = [5, 10, 5] #[5, 10, 5] 600ms, [8, 8, 4] 480ms
-param_dict = {"cache": dict(), "is_final": False, "chunk_size": chunk_size}
+chunk_size = [0, 10, 5] #[5, 10, 5] 600ms, [8, 8, 4] 480ms
+encoder_chunk_look_back = 4 #number of chunks to lookback for encoder self-attention
+decoder_chunk_look_back = 1 #number of encoder chunks to lookback for decoder cross-attention
+param_dict = {"cache": dict(), "is_final": False, "chunk_size": chunk_size,
+ "encoder_chunk_look_back": encoder_chunk_look_back, "decoder_chunk_look_back": decoder_chunk_look_back}
chunk_stride = chunk_size[1] * 960 # 600ms銆�480ms
# first chunk, 600ms
speech_chunk = speech[0:chunk_stride]
@@ -55,7 +58,7 @@
inference_pipeline = pipeline(
task=Tasks.auto_speech_recognition,
model='damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online',
- model_revision='v1.0.6',
+ model_revision='v1.0.7',
update_model=False,
mode="paraformer_fake_streaming"
)
diff --git a/egs_modelscope/asr/TEMPLATE/README_zh.md b/egs_modelscope/asr/TEMPLATE/README_zh.md
index 81e0271..6db310e 100644
--- a/egs_modelscope/asr/TEMPLATE/README_zh.md
+++ b/egs_modelscope/asr/TEMPLATE/README_zh.md
@@ -27,15 +27,18 @@
inference_pipeline = pipeline(
task=Tasks.auto_speech_recognition,
model='damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online',
- model_revision='v1.0.6',
+ model_revision='v1.0.7',
update_model=False,
mode='paraformer_streaming'
)
import soundfile
speech, sample_rate = soundfile.read("example/asr_example.wav")
-chunk_size = [5, 10, 5] #[5, 10, 5] 600ms, [8, 8, 4] 480ms
-param_dict = {"cache": dict(), "is_final": False, "chunk_size": chunk_size}
+chunk_size = [0, 10, 5] #[0, 10, 5] 600ms, [0, 8, 4] 480ms
+encoder_chunk_look_back = 4 #number of chunks to lookback for encoder self-attention
+decoder_chunk_look_back = 1 #number of encoder chunks to lookback for decoder cross-attention
+param_dict = {"cache": dict(), "is_final": False, "chunk_size": chunk_size,
+ "encoder_chunk_look_back": encoder_chunk_look_back, "decoder_chunk_look_back": decoder_chunk_look_back}
chunk_stride = chunk_size[1] * 960 # 600ms銆�480ms
# first chunk, 600ms
speech_chunk = speech[0:chunk_stride]
@@ -55,7 +58,7 @@
inference_pipeline = pipeline(
task=Tasks.auto_speech_recognition,
model='damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online',
- model_revision='v1.0.6',
+ model_revision='v1.0.7',
update_model=False,
mode="paraformer_fake_streaming"
)
--
Gitblit v1.9.1