from funasr import AutoModel
from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks
from modelscope import snapshot_download

# local_dir_root = "/models_from_modelscope"
# model_dir = snapshot_download('iic/speech_frcrn_ans_cirm_16k')

# model = AutoModel(model="paraformer-zh", vad_model="fsmn-vad",  punc_model="ct-punc",
#                   # spk_model="cam++",
#                   )
# res = model.generate(input="test1.mp3")
#
# text = res[0]["text"]
#
# print(text)

inference_pipeline = pipeline(
    task = Tasks.auto_speech_recognition,
    model = 'E:/Py_api_test/ASR_api/models_from_modelscope/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch',
    vad_model = "E:/Py_api_test/ASR_api/models_from_modelscope/speech_fsmn_vad_zh-cn-16k-common-pytorch",
    punc_model = "E:/Py_api_test/ASR_api/models_from_modelscope/punc_ct-transformer_cn-en-common-vocab471067-large",
)

param_dict = {}
param_dict['use_timestamp'] = False

rec_result = inference_pipeline(input="output.mp3", params=param_dict)
# print(rec_result["text"])