| | |
|---|
| | | # -*- encoding: utf-8 -*- |
|---|
| | | # Copyright FunASR (https://github.com/alibaba-damo-academy/FunASR). All Rights Reserved. |
|---|
| | | # MIT License (https://opensource.org/licenses/MIT) |
|---|
| | | |
|---|
| | | import os.path |
|---|
| | | from pathlib import Path |
|---|
| | |
|---|
| | | |
|---|
| | | |
|---|
| | | class Paraformer(): |
|---|
| | | """ |
|---|
| | | Author: Speech Lab of DAMO Academy, Alibaba Group |
|---|
| | | Paraformer: Fast and Accurate Parallel Transformer for Non-autoregressive End-to-End Speech Recognition |
|---|
| | | https://arxiv.org/abs/2206.08317 |
|---|
| | | """ |
|---|
| | | def __init__(self, model_dir: Union[str, Path] = None, |
|---|
| | | batch_size: int = 1, |
|---|
| | | device_id: Union[str, int] = "-1", |
|---|
| | | plot_timestamp_to: str = "", |
|---|
| | | pred_bias: int = 1, |
|---|
| | | quantize: bool = False, |
|---|
| | | intra_op_num_threads: int = 4, |
|---|
| | | ): |
|---|
| | |
|---|
| | | self.ort_infer = OrtInferSession(model_file, device_id, intra_op_num_threads=intra_op_num_threads) |
|---|
| | | self.batch_size = batch_size |
|---|
| | | self.plot_timestamp_to = plot_timestamp_to |
|---|
| | | self.pred_bias = pred_bias |
|---|
| | | if "predictor_bias" in config['model_conf'].keys(): |
|---|
| | | self.pred_bias = config['model_conf']['predictor_bias'] |
|---|
| | | else: |
|---|
| | | self.pred_bias = 0 |
|---|
| | | |
|---|
| | | def __call__(self, wav_content: Union[str, np.ndarray, List[str]], **kwargs) -> List: |
|---|
| | | waveform_list = self.load_data(wav_content, self.frontend.opts.frame_opts.samp_freq) |
|---|
