Skip to content

Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.

install requirements automatically

游雁447222c00e54
diff --git a/examples/common_voice/whisper_lid/demo_funasr.py b/examples/common_voice/whisper_lid/demo_funasr.py
index 9af790e1a..50f4e2ac2 100644
--- a/examples/common_voice/whisper_lid/demo_funasr.py
+++ b/examples/common_voice/whisper_lid/demo_funasr.py
@@ -12,7 +12,7 @@ multilingual_wavs = [
"example_ko.mp3",
]
-model = AutoModel(model="iic/speech_whisper-large_lid_multilingual_pytorch", model_revision="v2.0.4")
+model = AutoModel(model="iic/speech_whisper-large_lid_multilingual_pytorch", model_revision="master")
for wav_id in multilingual_wavs:
wav_file = f"{model.model_path}/examples/{wav_id}"
res = model.generate(input=wav_file, data_type="sound", inference_clip_length=250)
diff --git a/examples/common_voice/whisper_lid/demo_modelscope.py b/examples/common_voice/whisper_lid/demo_modelscope.py
index cce389a1c..e55a9720c 100644
--- a/examples/common_voice/whisper_lid/demo_modelscope.py
+++ b/examples/common_voice/whisper_lid/demo_modelscope.py
@@ -15,7 +15,7 @@ multilingual_wavs=[
inference_pipeline = pipeline(
task=Tasks.auto_speech_recognition,
- model='iic/speech_whisper-large_lid_multilingual_pytorch', model_revision="v2.0.4")
+ model='iic/speech_whisper-large_lid_multilingual_pytorch', model_revision="master")
for wav in multilingual_wavs:
rec_result = inference_pipeline(input=wav, inference_clip_length=250)
diff --git a/examples/industrial_data_pretraining/bicif_paraformer/demo.py b/examples/industrial_data_pretraining/bicif_paraformer/demo.py
index c8c4fa784..0b170659a 100644
--- a/examples/industrial_data_pretraining/bicif_paraformer/demo.py
+++ b/examples/industrial_data_pretraining/bicif_paraformer/demo.py
@@ -6,11 +6,11 @@
from funasr import AutoModel
model = AutoModel(model="iic/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch",
- model_revision="v2.0.4",
+ model_revision="master",
vad_model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch",
- vad_model_revision="v2.0.4",
+ vad_model_revision="master",
punc_model="iic/punc_ct-transformer_cn-en-common-vocab471067-large",
- punc_model_revision="v2.0.4",
+ punc_model_revision="master",
# spk_model="iic/speech_campplus_sv_zh-cn_16k-common",
# spk_model_revision="v2.0.2",
)
diff --git a/examples/industrial_data_pretraining/bicif_paraformer/demo.sh b/examples/industrial_data_pretraining/bicif_paraformer/demo.sh
index b6c1d1723..fe4473460 100644
--- a/examples/industrial_data_pretraining/bicif_paraformer/demo.sh
+++ b/examples/industrial_data_pretraining/bicif_paraformer/demo.sh
@@ -1,11 +1,11 @@
model="iic/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
-model_revision="v2.0.4"
+model_revision="master"
vad_model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch"
-vad_model_revision="v2.0.4"
+vad_model_revision="master"
#punc_model="iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch"
punc_model="iic/punc_ct-transformer_cn-en-common-vocab471067-large"
-punc_model_revision="v2.0.4"
+punc_model_revision="master"
spk_model="iic/speech_campplus_sv_zh-cn_16k-common"
spk_model_revision="v2.0.2"
diff --git a/examples/industrial_data_pretraining/bicif_paraformer/export.py b/examples/industrial_data_pretraining/bicif_paraformer/export.py
index 28a9e7b0e..8e45a2319 100644
--- a/examples/industrial_data_pretraining/bicif_paraformer/export.py
+++ b/examples/industrial_data_pretraining/bicif_paraformer/export.py
@@ -8,7 +8,7 @@
from funasr import AutoModel
model = AutoModel(model="iic/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch",
- model_revision="v2.0.4", device="cpu")
+ model_revision="master", device="cpu")
res = model.export(type="onnx", quantize=False)
print(res)
diff --git a/examples/industrial_data_pretraining/bicif_paraformer/export.sh b/examples/industrial_data_pretraining/bicif_paraformer/export.sh
index b6883b7c9..cf040eccc 100644
--- a/examples/industrial_data_pretraining/bicif_paraformer/export.sh
+++ b/examples/industrial_data_pretraining/bicif_paraformer/export.sh
@@ -6,7 +6,7 @@ export HYDRA_FULL_ERROR=1
model="iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
-model_revision="v2.0.4"
+model_revision="master"
python -m funasr.bin.export \
++model=${model} \
diff --git a/examples/industrial_data_pretraining/bicif_paraformer/finetune.sh b/examples/industrial_data_pretraining/bicif_paraformer/finetune.sh
index 885e5c0e2..9d251a107 100644
--- a/examples/industrial_data_pretraining/bicif_paraformer/finetune.sh
+++ b/examples/industrial_data_pretraining/bicif_paraformer/finetune.sh
@@ -11,7 +11,7 @@ gpu_num=$(echo $CUDA_VISIBLE_DEVICES | awk -F "," '{print NF}')
## option 1, download model automatically
model_name_or_model_dir="iic/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
-model_revision="v2.0.4"
+model_revision="master"
## option 2, download model by git
#local_path_root=${workspace}/modelscope_models
diff --git a/examples/industrial_data_pretraining/conformer/demo.py b/examples/industrial_data_pretraining/conformer/demo.py
index 358a1f800..1abc7a7da 100644
--- a/examples/industrial_data_pretraining/conformer/demo.py
+++ b/examples/industrial_data_pretraining/conformer/demo.py
@@ -5,7 +5,7 @@
from funasr import AutoModel
-model = AutoModel(model="iic/speech_conformer_asr_nat-zh-cn-16k-aishell2-vocab5212-pytorch", model_revision="v2.0.4",
+model = AutoModel(model="iic/speech_conformer_asr_nat-zh-cn-16k-aishell2-vocab5212-pytorch", model_revision="master",
)
res = model.generate(input="https://modelscope.oss-cn-beijing.aliyuncs.com/test/audios/asr_example.wav")
diff --git a/examples/industrial_data_pretraining/conformer/demo.sh b/examples/industrial_data_pretraining/conformer/demo.sh
index c259799f3..9cf6cc535 100644
--- a/examples/industrial_data_pretraining/conformer/demo.sh
+++ b/examples/industrial_data_pretraining/conformer/demo.sh
@@ -1,6 +1,6 @@
model="iic/speech_conformer_asr_nat-zh-cn-16k-aishell1-vocab4234-pytorch"
-model_revision="v2.0.4"
+model_revision="master"
python funasr/bin/inference.py \
+model=${model} \
diff --git a/examples/industrial_data_pretraining/contextual_paraformer/demo.py b/examples/industrial_data_pretraining/contextual_paraformer/demo.py
index 4168a7d73..cd93a4fce 100755
--- a/examples/industrial_data_pretraining/contextual_paraformer/demo.py
+++ b/examples/industrial_data_pretraining/contextual_paraformer/demo.py
@@ -5,7 +5,7 @@
from funasr import AutoModel
-model = AutoModel(model="iic/speech_paraformer-large-contextual_asr_nat-zh-cn-16k-common-vocab8404", model_revision="v2.0.4")
+model = AutoModel(model="iic/speech_paraformer-large-contextual_asr_nat-zh-cn-16k-common-vocab8404", model_revision="master")
res = model.generate(input="https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_zh.wav",
hotword='达摩院 魔搭')
diff --git a/examples/industrial_data_pretraining/contextual_paraformer/demo.sh b/examples/industrial_data_pretraining/contextual_paraformer/demo.sh
index 987f2ff29..e6513388f 100755
--- a/examples/industrial_data_pretraining/contextual_paraformer/demo.sh
+++ b/examples/industrial_data_pretraining/contextual_paraformer/demo.sh
@@ -1,6 +1,6 @@
model="iic/speech_paraformer-large-contextual_asr_nat-zh-cn-16k-common-vocab8404"
-model_revision="v2.0.4"
+model_revision="master"
python ../../../funasr/bin/inference.py \
+model=${model} \
diff --git a/examples/industrial_data_pretraining/contextual_paraformer/finetune.sh b/examples/industrial_data_pretraining/contextual_paraformer/finetune.sh
index de8896853..fe12315bc 100644
--- a/examples/industrial_data_pretraining/contextual_paraformer/finetune.sh
+++ b/examples/industrial_data_pretraining/contextual_paraformer/finetune.sh
@@ -11,7 +11,7 @@ gpu_num=$(echo $CUDA_VISIBLE_DEVICES | awk -F "," '{print NF}')
## option 1, download model automatically
model_name_or_model_dir="iic/speech_paraformer-large-contextual_asr_nat-zh-cn-16k-common-vocab8404"
-model_revision="v2.0.4"
+model_revision="master"
## option 2, download model by git
#local_path_root=${workspace}/modelscope_models
diff --git a/examples/industrial_data_pretraining/ct_transformer/demo.py b/examples/industrial_data_pretraining/ct_transformer/demo.py
index 31755dcc1..6e6b47809 100644
--- a/examples/industrial_data_pretraining/ct_transformer/demo.py
+++ b/examples/industrial_data_pretraining/ct_transformer/demo.py
@@ -5,7 +5,7 @@
from funasr import AutoModel
-model = AutoModel(model="iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch", model_revision="v2.0.4")
+model = AutoModel(model="iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch", model_revision="master")
res = model.generate(input="https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_text/punc_example.txt")
print(res)
@@ -13,7 +13,7 @@ print(res)
from funasr import AutoModel
-model = AutoModel(model="iic/punc_ct-transformer_cn-en-common-vocab471067-large", model_revision="v2.0.4")
+model = AutoModel(model="iic/punc_ct-transformer_cn-en-common-vocab471067-large", model_revision="master")
res = model.generate(input="https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_text/punc_example.txt")
print(res)
\ No newline at end of file
diff --git a/examples/industrial_data_pretraining/ct_transformer/demo.sh b/examples/industrial_data_pretraining/ct_transformer/demo.sh
index 56d0bd378..02ee5a8d9 100644
--- a/examples/industrial_data_pretraining/ct_transformer/demo.sh
+++ b/examples/industrial_data_pretraining/ct_transformer/demo.sh
@@ -1,9 +1,9 @@
#model="iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch"
-#model_revision="v2.0.4"
+#model_revision="master"
model="iic/punc_ct-transformer_cn-en-common-vocab471067-large"
-model_revision="v2.0.4"
+model_revision="master"
python funasr/bin/inference.py \
+model=${model} \
diff --git a/examples/industrial_data_pretraining/ct_transformer/export.py b/examples/industrial_data_pretraining/ct_transformer/export.py
index 8c35670be..397bb9691 100644
--- a/examples/industrial_data_pretraining/ct_transformer/export.py
+++ b/examples/industrial_data_pretraining/ct_transformer/export.py
@@ -8,7 +8,7 @@
from funasr import AutoModel
model = AutoModel(model="iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch",
- model_revision="v2.0.4")
+ model_revision="master")
res = model.export(type="onnx", quantize=False)
print(res)
diff --git a/examples/industrial_data_pretraining/ct_transformer/export.sh b/examples/industrial_data_pretraining/ct_transformer/export.sh
index a11cda5d4..5f7e4fb87 100644
--- a/examples/industrial_data_pretraining/ct_transformer/export.sh
+++ b/examples/industrial_data_pretraining/ct_transformer/export.sh
@@ -6,7 +6,7 @@ export HYDRA_FULL_ERROR=1
model="iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch"
-model_revision="v2.0.4"
+model_revision="master"
python -m funasr.bin.export \
++model=${model} \
diff --git a/examples/industrial_data_pretraining/ct_transformer_streaming/demo.py b/examples/industrial_data_pretraining/ct_transformer_streaming/demo.py
index 66cc5c5b8..14edbe4b5 100644
--- a/examples/industrial_data_pretraining/ct_transformer_streaming/demo.py
+++ b/examples/industrial_data_pretraining/ct_transformer_streaming/demo.py
@@ -5,7 +5,7 @@
from funasr import AutoModel
-model = AutoModel(model="iic/punc_ct-transformer_zh-cn-common-vad_realtime-vocab272727", model_revision="v2.0.4")
+model = AutoModel(model="iic/punc_ct-transformer_zh-cn-common-vad_realtime-vocab272727", model_revision="master")
inputs = "跨境河流是养育沿岸|人民的生命之源长期以来为帮助下游地区防灾减灾中方技术人员|在上游地区极为恶劣的自然条件下克服巨大困难甚至冒着生命危险|向印方提供汛期水文资料处理紧急事件中方重视印方在跨境河流问题上的关切|愿意进一步完善双方联合工作机制|凡是|中方能做的我们|都会去做而且会做得更好我请印度朋友们放心中国在上游的|任何开发利用都会经过科学|规划和论证兼顾上下游的利益"
vads = inputs.split("|")
diff --git a/examples/industrial_data_pretraining/ct_transformer_streaming/demo.sh b/examples/industrial_data_pretraining/ct_transformer_streaming/demo.sh
index 7124fd565..ad3e3eac4 100644
--- a/examples/industrial_data_pretraining/ct_transformer_streaming/demo.sh
+++ b/examples/industrial_data_pretraining/ct_transformer_streaming/demo.sh
@@ -1,6 +1,6 @@
model="iic/punc_ct-transformer_zh-cn-common-vad_realtime-vocab272727"
-model_revision="v2.0.4"
+model_revision="master"
python funasr/bin/inference.py \
+model=${model} \
diff --git a/examples/industrial_data_pretraining/ct_transformer_streaming/export.py b/examples/industrial_data_pretraining/ct_transformer_streaming/export.py
index 47fa08a89..2e3b17276 100644
--- a/examples/industrial_data_pretraining/ct_transformer_streaming/export.py
+++ b/examples/industrial_data_pretraining/ct_transformer_streaming/export.py
@@ -8,7 +8,7 @@
from funasr import AutoModel
model = AutoModel(model="iic/punc_ct-transformer_zh-cn-common-vad_realtime-vocab272727",
- model_revision="v2.0.4")
+ model_revision="master")
res = model.export(type="onnx", quantize=False)
print(res)
diff --git a/examples/industrial_data_pretraining/ct_transformer_streaming/export.sh b/examples/industrial_data_pretraining/ct_transformer_streaming/export.sh
index 118afbbb6..a47f70157 100644
--- a/examples/industrial_data_pretraining/ct_transformer_streaming/export.sh
+++ b/examples/industrial_data_pretraining/ct_transformer_streaming/export.sh
@@ -6,7 +6,7 @@ export HYDRA_FULL_ERROR=1
model="iic/punc_ct-transformer_zh-cn-common-vad_realtime-vocab272727"
-model_revision="v2.0.4"
+model_revision="master"
python -m funasr.bin.export \
++model=${model} \
diff --git a/examples/industrial_data_pretraining/emotion2vec/demo.py b/examples/industrial_data_pretraining/emotion2vec/demo.py
index 510e6e861..50deda3cd 100644
--- a/examples/industrial_data_pretraining/emotion2vec/demo.py
+++ b/examples/industrial_data_pretraining/emotion2vec/demo.py
@@ -6,9 +6,9 @@
from funasr import AutoModel
# model="iic/emotion2vec_base"
-model = AutoModel(model="iic/emotion2vec_base_finetuned", model_revision="v2.0.4",
+model = AutoModel(model="iic/emotion2vec_base_finetuned", model_revision="master",
# vad_model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch",
- # vad_model_revision="v2.0.4",
+ # vad_model_revision="master",
# vad_kwargs={"max_single_segment_time": 2000},
)
diff --git a/examples/industrial_data_pretraining/fsmn_vad_streaming/demo.py b/examples/industrial_data_pretraining/fsmn_vad_streaming/demo.py
index 03c01ea06..61dce49be 100644
--- a/examples/industrial_data_pretraining/fsmn_vad_streaming/demo.py
+++ b/examples/industrial_data_pretraining/fsmn_vad_streaming/demo.py
@@ -6,7 +6,7 @@
from funasr import AutoModel
wav_file = "https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/vad_example.wav"
-model = AutoModel(model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch", model_revision="v2.0.4")
+model = AutoModel(model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch", model_revision="master")
res = model.generate(input=wav_file)
print(res)
diff --git a/examples/industrial_data_pretraining/fsmn_vad_streaming/demo.sh b/examples/industrial_data_pretraining/fsmn_vad_streaming/demo.sh
index 8cdf3edf6..0248dd03e 100644
--- a/examples/industrial_data_pretraining/fsmn_vad_streaming/demo.sh
+++ b/examples/industrial_data_pretraining/fsmn_vad_streaming/demo.sh
@@ -1,7 +1,7 @@
model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch"
-model_revision="v2.0.4"
+model_revision="master"
python funasr/bin/inference.py \
+model=${model} \
diff --git a/examples/industrial_data_pretraining/fsmn_vad_streaming/export.py b/examples/industrial_data_pretraining/fsmn_vad_streaming/export.py
index 2c8fd4d6d..f45ddeebc 100644
--- a/examples/industrial_data_pretraining/fsmn_vad_streaming/export.py
+++ b/examples/industrial_data_pretraining/fsmn_vad_streaming/export.py
@@ -8,7 +8,7 @@
from funasr import AutoModel
-model = AutoModel(model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch", model_revision="v2.0.4")
+model = AutoModel(model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch", model_revision="master")
res = model.export(type="onnx", quantize=False)
print(res)
diff --git a/examples/industrial_data_pretraining/fsmn_vad_streaming/export.sh b/examples/industrial_data_pretraining/fsmn_vad_streaming/export.sh
index 1a8207a69..9143dc3db 100644
--- a/examples/industrial_data_pretraining/fsmn_vad_streaming/export.sh
+++ b/examples/industrial_data_pretraining/fsmn_vad_streaming/export.sh
@@ -7,7 +7,7 @@ export HYDRA_FULL_ERROR=1
model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch"
-model_revision="v2.0.4"
+model_revision="master"
python -m funasr.bin.export \
++model=${model} \
diff --git a/examples/industrial_data_pretraining/monotonic_aligner/demo.py b/examples/industrial_data_pretraining/monotonic_aligner/demo.py
index 2e76f649f..467de1bbc 100644
--- a/examples/industrial_data_pretraining/monotonic_aligner/demo.py
+++ b/examples/industrial_data_pretraining/monotonic_aligner/demo.py
@@ -5,7 +5,7 @@
from funasr import AutoModel
-model = AutoModel(model="iic/speech_timestamp_prediction-v1-16k-offline", model_revision="v2.0.4")
+model = AutoModel(model="iic/speech_timestamp_prediction-v1-16k-offline", model_revision="master")
res = model.generate(input=("https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_zh.wav",
"欢迎大家来到魔搭社区进行体验"),
diff --git a/examples/industrial_data_pretraining/monotonic_aligner/demo.sh b/examples/industrial_data_pretraining/monotonic_aligner/demo.sh
index ece97ff76..649ce7009 100644
--- a/examples/industrial_data_pretraining/monotonic_aligner/demo.sh
+++ b/examples/industrial_data_pretraining/monotonic_aligner/demo.sh
@@ -1,6 +1,6 @@
model="iic/speech_timestamp_prediction-v1-16k-offline"
-model_revision="v2.0.4"
+model_revision="master"
python funasr/bin/inference.py \
+model=${model} \
diff --git a/examples/industrial_data_pretraining/paraformer-zh-spk/demo.py b/examples/industrial_data_pretraining/paraformer-zh-spk/demo.py
index 6ed021713..523bb3a24 100644
--- a/examples/industrial_data_pretraining/paraformer-zh-spk/demo.py
+++ b/examples/industrial_data_pretraining/paraformer-zh-spk/demo.py
@@ -6,11 +6,11 @@
from funasr import AutoModel
model = AutoModel(model="iic/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch",
- model_revision="v2.0.4",
+ model_revision="master",
vad_model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch",
- vad_model_revision="v2.0.4",
+ vad_model_revision="master",
punc_model="iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch",
- punc_model_revision="v2.0.4",
+ punc_model_revision="master",
# spk_model="iic/speech_campplus_sv_zh-cn_16k-common",
# spk_model_revision="v2.0.2"
)
diff --git a/examples/industrial_data_pretraining/paraformer-zh-spk/demo.sh b/examples/industrial_data_pretraining/paraformer-zh-spk/demo.sh
index dae1aca6f..fb0c4d902 100644
--- a/examples/industrial_data_pretraining/paraformer-zh-spk/demo.sh
+++ b/examples/industrial_data_pretraining/paraformer-zh-spk/demo.sh
@@ -1,10 +1,10 @@
model="iic/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
-model_revision="v2.0.4"
+model_revision="master"
vad_model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch"
-vad_model_revision="v2.0.4"
+vad_model_revision="master"
punc_model="iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch"
-punc_model_revision="v2.0.4"
+punc_model_revision="master"
spk_model="iic/speech_campplus_sv_zh-cn_16k-common"
spk_model_revision="v2.0.2"
diff --git a/examples/industrial_data_pretraining/paraformer/demo.py b/examples/industrial_data_pretraining/paraformer/demo.py
index a39a52673..6cef23442 100644
--- a/examples/industrial_data_pretraining/paraformer/demo.py
+++ b/examples/industrial_data_pretraining/paraformer/demo.py
@@ -6,12 +6,12 @@
from funasr import AutoModel
model = AutoModel(model="iic/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch",
- model_revision="v2.0.4",
+ model_revision="master",
vad_model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch",
- vad_model_revision="v2.0.4",
+ vad_model_revision="master",
vad_kwargs={"max_single_segment_time": 60000},
punc_model="iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch",
- punc_model_revision="v2.0.4",
+ punc_model_revision="master",
# spk_model="iic/speech_campplus_sv_zh-cn_16k-common",
# spk_model_revision="v2.0.2",
)
@@ -23,7 +23,7 @@ print(res)
''' can not use currently
from funasr import AutoFrontend
-frontend = AutoFrontend(model="iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch", model_revision="v2.0.4")
+frontend = AutoFrontend(model="iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch", model_revision="master")
fbanks = frontend(input="https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_zh.wav", batch_size=2)
diff --git a/examples/industrial_data_pretraining/paraformer/export.py b/examples/industrial_data_pretraining/paraformer/export.py
index fce4d7712..0c181c170 100644
--- a/examples/industrial_data_pretraining/paraformer/export.py
+++ b/examples/industrial_data_pretraining/paraformer/export.py
@@ -10,7 +10,7 @@
from funasr import AutoModel
model = AutoModel(model="iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch",
- model_revision="v2.0.4")
+ model_revision="master")
res = model.export(type="onnx", quantize=False)
print(res)
diff --git a/examples/industrial_data_pretraining/paraformer/export.sh b/examples/industrial_data_pretraining/paraformer/export.sh
index fc341e74e..18d2039c5 100644
--- a/examples/industrial_data_pretraining/paraformer/export.sh
+++ b/examples/industrial_data_pretraining/paraformer/export.sh
@@ -6,7 +6,7 @@ export HYDRA_FULL_ERROR=1
model="iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
-model_revision="v2.0.4"
+model_revision="master"
python -m funasr.bin.export \
diff --git a/examples/industrial_data_pretraining/paraformer/finetune.sh b/examples/industrial_data_pretraining/paraformer/finetune.sh
index 3041ff399..408076b56 100644
--- a/examples/industrial_data_pretraining/paraformer/finetune.sh
+++ b/examples/industrial_data_pretraining/paraformer/finetune.sh
@@ -10,7 +10,7 @@ gpu_num=$(echo $CUDA_VISIBLE_DEVICES | awk -F "," '{print NF}')
## option 1, download model automatically
model_name_or_model_dir="iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
-model_revision="v2.0.4"
+model_revision="master"
## option 2, download model by git
#local_path_root=${workspace}/modelscope_models
diff --git a/examples/industrial_data_pretraining/paraformer/infer.sh b/examples/industrial_data_pretraining/paraformer/infer.sh
index aa523e5a5..0b0f931ab 100644
--- a/examples/industrial_data_pretraining/paraformer/infer.sh
+++ b/examples/industrial_data_pretraining/paraformer/infer.sh
@@ -9,7 +9,7 @@ input="https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr
output_dir="./outputs/debug"
model="iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
-model_revision="v2.0.4"
+model_revision="master"
device="cuda:0" # "cuda:0" for gpu0, "cuda:1" for gpu1, "cpu"
diff --git a/examples/industrial_data_pretraining/paraformer_streaming/demo.py b/examples/industrial_data_pretraining/paraformer_streaming/demo.py
index 9fcee3a8c..94e52008c 100644
--- a/examples/industrial_data_pretraining/paraformer_streaming/demo.py
+++ b/examples/industrial_data_pretraining/paraformer_streaming/demo.py
@@ -10,7 +10,7 @@ from funasr import AutoModel
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
-model = AutoModel(model="iic/speech_paraformer_asr_nat-zh-cn-16k-common-vocab8404-online", model_revision="v2.0.4")
+model = AutoModel(model="iic/speech_paraformer_asr_nat-zh-cn-16k-common-vocab8404-online", model_revision="master")
wav_file = os.path.join(model.model_path, "example/asr_example.wav")
res = model.generate(input=wav_file,
diff --git a/examples/industrial_data_pretraining/paraformer_streaming/demo.sh b/examples/industrial_data_pretraining/paraformer_streaming/demo.sh
index c3f7bb488..a316aafe8 100644
--- a/examples/industrial_data_pretraining/paraformer_streaming/demo.sh
+++ b/examples/industrial_data_pretraining/paraformer_streaming/demo.sh
@@ -1,6 +1,6 @@
model="iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online"
-model_revision="v2.0.4"
+model_revision="master"
python funasr/bin/inference.py \
+model=${model} \
diff --git a/examples/industrial_data_pretraining/paraformer_streaming/export.py b/examples/industrial_data_pretraining/paraformer_streaming/export.py
index 8e223108b..06cec31ac 100644
--- a/examples/industrial_data_pretraining/paraformer_streaming/export.py
+++ b/examples/industrial_data_pretraining/paraformer_streaming/export.py
@@ -10,7 +10,7 @@
from funasr import AutoModel
model = AutoModel(model="iic/speech_paraformer_asr_nat-zh-cn-16k-common-vocab8404-online",
- model_revision="v2.0.4")
+ model_revision="master")
res = model.export(type="onnx", quantize=False)
print(res)
diff --git a/examples/industrial_data_pretraining/paraformer_streaming/export.sh b/examples/industrial_data_pretraining/paraformer_streaming/export.sh
index 43e344b45..25ac5139c 100644
--- a/examples/industrial_data_pretraining/paraformer_streaming/export.sh
+++ b/examples/industrial_data_pretraining/paraformer_streaming/export.sh
@@ -6,7 +6,7 @@ export HYDRA_FULL_ERROR=1
model="iic/speech_paraformer_asr_nat-zh-cn-16k-common-vocab8404-online"
-model_revision="v2.0.4"
+model_revision="master"
python -m funasr.bin.export \
diff --git a/examples/industrial_data_pretraining/paraformer_streaming/finetune.sh b/examples/industrial_data_pretraining/paraformer_streaming/finetune.sh
index 2bfb2bf39..1e4537a7b 100644
--- a/examples/industrial_data_pretraining/paraformer_streaming/finetune.sh
+++ b/examples/industrial_data_pretraining/paraformer_streaming/finetune.sh
@@ -10,7 +10,7 @@ gpu_num=$(echo $CUDA_VISIBLE_DEVICES | awk -F "," '{print NF}')
## option 1, download model automatically
model_name_or_model_dir="iic/speech_paraformer_asr_nat-zh-cn-16k-common-vocab8404-online"
-model_revision="v2.0.4"
+model_revision="master"
## option 2, download model by git
#local_path_root=${workspace}/modelscope_models
diff --git a/examples/industrial_data_pretraining/scama/demo.py b/examples/industrial_data_pretraining/scama/demo.py
index 19996233a..075039a38 100644
--- a/examples/industrial_data_pretraining/scama/demo.py
+++ b/examples/industrial_data_pretraining/scama/demo.py
@@ -9,7 +9,7 @@ chunk_size = [5, 10, 5] #[0, 10, 5] 600ms, [0, 8, 4] 480ms
encoder_chunk_look_back = 0 #number of chunks to lookback for encoder self-attention
decoder_chunk_look_back = 0 #number of encoder chunks to lookback for decoder cross-attention
-model = AutoModel(model="/Users/zhifu/Downloads/modelscope_models/speech_SCAMA_asr-zh-cn-16k-common-vocab8358-streaming", model_revision="v2.0.4")
+model = AutoModel(model="/Users/zhifu/Downloads/modelscope_models/speech_SCAMA_asr-zh-cn-16k-common-vocab8358-streaming", model_revision="master")
cache = {}
res = model.generate(input="https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_zh.wav",
chunk_size=chunk_size,
diff --git a/examples/industrial_data_pretraining/scama/demo.sh b/examples/industrial_data_pretraining/scama/demo.sh
index c3f7bb488..a316aafe8 100644
--- a/examples/industrial_data_pretraining/scama/demo.sh
+++ b/examples/industrial_data_pretraining/scama/demo.sh
@@ -1,6 +1,6 @@
model="iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online"
-model_revision="v2.0.4"
+model_revision="master"
python funasr/bin/inference.py \
+model=${model} \
diff --git a/examples/industrial_data_pretraining/seaco_paraformer/demo.py b/examples/industrial_data_pretraining/seaco_paraformer/demo.py
index 69e9020f4..c12b279bc 100644
--- a/examples/industrial_data_pretraining/seaco_paraformer/demo.py
+++ b/examples/industrial_data_pretraining/seaco_paraformer/demo.py
@@ -6,11 +6,11 @@
from funasr import AutoModel
model = AutoModel(model="iic/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch",
- model_revision="v2.0.4",
+ model_revision="master",
# vad_model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch",
- # vad_model_revision="v2.0.4",
+ # vad_model_revision="master",
# punc_model="iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch",
- # punc_model_revision="v2.0.4",
+ # punc_model_revision="master",
# spk_model="iic/speech_campplus_sv_zh-cn_16k-common",
# spk_model_revision="v2.0.2",
)
diff --git a/examples/industrial_data_pretraining/seaco_paraformer/demo.sh b/examples/industrial_data_pretraining/seaco_paraformer/demo.sh
index 9a02572fa..65847c7f4 100644
--- a/examples/industrial_data_pretraining/seaco_paraformer/demo.sh
+++ b/examples/industrial_data_pretraining/seaco_paraformer/demo.sh
@@ -1,10 +1,10 @@
model="iic/speech_seaco_paraformer_large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
-model_revision="v2.0.4"
+model_revision="master"
vad_model="iic/speech_fsmn_vad_zh-cn-16k-common-pytorch"
-vad_model_revision="v2.0.4"
+vad_model_revision="master"
punc_model="iic/punc_ct-transformer_zh-cn-common-vocab272727-pytorch"
-punc_model_revision="v2.0.4"
+punc_model_revision="master"
python funasr/bin/inference.py \
+model=${model} \
diff --git a/examples/industrial_data_pretraining/uniasr/demo.py b/examples/industrial_data_pretraining/uniasr/demo.py
index 3006d5eae..0df06e27c 100644
--- a/examples/industrial_data_pretraining/uniasr/demo.py
+++ b/examples/industrial_data_pretraining/uniasr/demo.py
@@ -6,7 +6,7 @@
from funasr import AutoModel
-model = AutoModel(model="iic/speech_UniASR-large_asr_2pass-zh-cn-16k-common-vocab8358-tensorflow1-offline", model_revision="v2.0.4",)
+model = AutoModel(model="iic/speech_UniASR-large_asr_2pass-zh-cn-16k-common-vocab8358-tensorflow1-offline", model_revision="master",)
res = model.generate(input="https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_zh.wav")
@@ -16,7 +16,7 @@ print(res)
''' can not use currently
from funasr import AutoFrontend
-frontend = AutoFrontend(model="iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch", model_revision="v2.0.4")
+frontend = AutoFrontend(model="iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch", model_revision="master")
fbanks = frontend(input="https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_zh.wav", batch_size=2)
diff --git a/examples/industrial_data_pretraining/uniasr/demo.sh b/examples/industrial_data_pretraining/uniasr/demo.sh
index 71f249ac3..e38f97419 100644
--- a/examples/industrial_data_pretraining/uniasr/demo.sh
+++ b/examples/industrial_data_pretraining/uniasr/demo.sh
@@ -1,6 +1,6 @@
model="iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
-model_revision="v2.0.4"
+model_revision="master"
python funasr/bin/inference.py \
+model=${model} \
diff --git a/examples/industrial_data_pretraining/whisper/infer.sh b/examples/industrial_data_pretraining/whisper/infer.sh
index 5beb7e2d5..6e8f2479b 100644
--- a/examples/industrial_data_pretraining/whisper/infer.sh
+++ b/examples/industrial_data_pretraining/whisper/infer.sh
@@ -11,7 +11,7 @@ input="https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr
output_dir="./outputs/debug"
model="iic/speech_whisper-large_asr_multilingual"
-model_revision="v2.0.4"
+model_revision="master"
device="cuda:0" # "cuda:0" for gpu0, "cuda:1" for gpu1, "cpu"
diff --git a/funasr/auto/auto_model.py b/funasr/auto/auto_model.py
index c31308e33..73ddc199c 100644
--- a/funasr/auto/auto_model.py
+++ b/funasr/auto/auto_model.py
@@ -109,7 +109,7 @@ class AutoModel:
if vad_model is not None:
logging.info("Building VAD model.")
vad_kwargs["model"] = vad_model
- vad_kwargs["model_revision"] = kwargs.get("vad_model_revision", None)
+ vad_kwargs["model_revision"] = kwargs.get("vad_model_revision", "master")
vad_kwargs["device"] = kwargs["device"]
vad_model, vad_kwargs = self.build_model(**vad_kwargs)
@@ -119,7 +119,7 @@ class AutoModel:
if punc_model is not None:
logging.info("Building punc model.")
punc_kwargs["model"] = punc_model
- punc_kwargs["model_revision"] = kwargs.get("punc_model_revision", None)
+ punc_kwargs["model_revision"] = kwargs.get("punc_model_revision", "master")
punc_kwargs["device"] = kwargs["device"]
punc_model, punc_kwargs = self.build_model(**punc_kwargs)
@@ -129,7 +129,7 @@ class AutoModel:
if spk_model is not None:
logging.info("Building SPK model.")
spk_kwargs["model"] = spk_model
- spk_kwargs["model_revision"] = kwargs.get("spk_model_revision", None)
+ spk_kwargs["model_revision"] = kwargs.get("spk_model_revision", "master")
spk_kwargs["device"] = kwargs["device"]
spk_model, spk_kwargs = self.build_model(**spk_kwargs)
self.cb_model = ClusterBackend().to(kwargs["device"])
diff --git a/funasr/download/download_from_hub.py b/funasr/download/download_from_hub.py
index 46704b096..75915f55e 100644
--- a/funasr/download/download_from_hub.py
+++ b/funasr/download/download_from_hub.py
@@ -29,7 +29,7 @@ def download_from_ms(**kwargs):
model_or_path = kwargs.get("model")
if model_or_path in name_maps_ms:
model_or_path = name_maps_ms[model_or_path]
- model_revision = kwargs.get("model_revision")
+ model_revision = kwargs.get("model_revision", "master")
if not os.path.exists(model_or_path) and "model_path" not in kwargs:
try:
model_or_path = get_or_download_model_dir(model_or_path, model_revision,
diff --git a/funasr/models/llm_asr/model.py b/funasr/models/llm_asr/model.py
index f5dd3b108..90cbd948a 100644
--- a/funasr/models/llm_asr/model.py
+++ b/funasr/models/llm_asr/model.py
@@ -73,7 +73,7 @@ class LLMASR(nn.Module):
hub = audio_encoder_conf.get("hub", None)
if hub == "ms":
from funasr import AutoModel
- model = AutoModel(model=audio_encoder, model_revision="v2.0.4")
+ model = AutoModel(model=audio_encoder, model_revision="master")
# frontend = model.kwargs.get("frontend")
audio_encoder_output_size = model.model.encoder_output_size
diff --git a/funasr/models/llm_asr_nar/model.py b/funasr/models/llm_asr_nar/model.py
index 06b21939f..30537cf79 100644
--- a/funasr/models/llm_asr_nar/model.py
+++ b/funasr/models/llm_asr_nar/model.py
@@ -75,7 +75,7 @@ class LLMASRNAR(nn.Module):
if hub == "funasr":
from funasr import AutoModel
init_param_path = encoder_conf.get("init_param_path", "iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch")
- model = AutoModel(model=init_param_path, model_revision="v2.0.4")
+ model = AutoModel(model=init_param_path, model_revision="master")
# frontend = model.kwargs.get("frontend")
model.model.decoder = None
@@ -406,7 +406,7 @@ class LLMASRNARPrompt(nn.Module):
from funasr import AutoModel
init_param_path = encoder_conf.get("init_param_path",
"iic/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch")
- model = AutoModel(model=init_param_path, model_revision="v2.0.4")
+ model = AutoModel(model=init_param_path, model_revision="master")
# frontend = model.kwargs.get("frontend")
model.model.decoder = None