Merge branch 'main' of github.com:alibaba-damo-academy/FunASR add

游雁

2023-01-17 241e049d0f7542fc91da18842a27f037c84a8c3c

Merge branch 'main' of github.com:alibaba-damo-academy/FunASR
add

 docs/get_started.md

@@ -110,7 +110,7 @@

* Configuration

We support CTC decoding, attention decoding and hybrid CTC-attention decoding in FunASR, which can be specified by `ctc_weight` in a YAML file in `conf` directory. Specifically, `ctc_weight=1.0` indicates attention decoding, `ctc_weight=0.0` indicates CTC decoding, `0.0<ctc_weight<1.0` indicates hybrid CTC-attention decoding.
We support CTC decoding, attention decoding and hybrid CTC-attention decoding in FunASR, which can be specified by `ctc_weight` in a YAML file in `conf` directory. Specifically, `ctc_weight=1.0` indicates CTC decoding, `ctc_weight=0.0` indicates attention decoding, `0.0<ctc_weight<1.0` indicates hybrid CTC-attention decoding.

* CPU/GPU Decoding


 egs_modelscope/asr/paraformer/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch/README.md

File was deleted

 egs_modelscope/asr/paraformer/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch/finetune.py

File was deleted

 egs_modelscope/asr/paraformer/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch/infer.py

File was deleted

 egs_modelscope/asr/paraformer/speech_paraformer-large_asr_nat-zh-cn-16k-aishell1-vocab8404-pytorch/infer.py

@@ -67,7 +67,7 @@
    for file in files:
        with open(os.path.join(best_recog_path, file), "w") as f:
            for i in range(nj):
                job_file = os.path.join(output_dir, "output.{}".format(str(i + 1)), file)
                job_file = os.path.join(output_dir, "output.{}/1best_recog".format(str(i + 1)), file)
                with open(job_file) as f_job:
                    lines = f_job.readlines()
                f.writelines(lines)

 egs_modelscope/asr/paraformer/speech_paraformer-large_asr_nat-zh-cn-16k-aishell2-vocab8404-pytorch/infer.py

@@ -67,7 +67,7 @@
    for file in files:
        with open(os.path.join(best_recog_path, file), "w") as f:
            for i in range(nj):
                job_file = os.path.join(output_dir, "output.{}".format(str(i + 1)), file)
                job_file = os.path.join(output_dir, "output.{}/1best_recog".format(str(i + 1)), file)
                with open(job_file) as f_job:
                    lines = f_job.readlines()
                f.writelines(lines)

 egs_modelscope/asr/paraformer/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch/infer.py

@@ -67,7 +67,7 @@
    for file in files:
        with open(os.path.join(best_recog_path, file), "w") as f:
            for i in range(nj):
                job_file = os.path.join(output_dir, "output.{}".format(str(i + 1)), file)
                job_file = os.path.join(output_dir, "output.{}/1best_recog".format(str(i + 1)), file)
                with open(job_file) as f_job:
                    lines = f_job.readlines()
                f.writelines(lines)

 egs_modelscope/asr/paraformer/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch/local_infer.py

File was deleted

 funasr/bin/asr_inference_paraformer.py

@@ -259,7 +259,7 @@
                    token_int = hyp.yseq[1:last_pos].tolist()

                # remove blank symbol id, which is assumed to be 0
                token_int = list(filter(lambda x: x != 0, token_int))
                token_int = list(filter(lambda x: x != 0 and x != 2, token_int))

                # Change integer-ids to tokens
                token = self.converter.ids2tokens(token_int)
@@ -650,7 +650,7 @@
                        finish_count += 1
                        # asr_utils.print_progress(finish_count / file_count)
                        if writer is not None:
                            ibest_writer["text"][key] = text
                            ibest_writer["text"][key] = text_postprocessed

                    logging.info("decoding, utt: {}, predictions: {}".format(key, text))
        rtf_avg = "decoding, feature length total: {}, forward_time total: {:.4f}, rtf avg: {:.4f}".format(length_total, forward_time_total, 100 * forward_time_total / (length_total * lfr_factor))