funasr1.0

游雁

2024-01-15 a035d68e860ea6decdf422c0fc04eda4fc4de397

 README_zh.md

@@ -119,13 +119,7 @@
for i in range(total_chunk_num):
    speech_chunk = speech[i*chunk_stride:(i+1)*chunk_stride]
    is_final = i == total_chunk_num - 1
    res = model(input=speech_chunk,
                cache=cache,
                is_final=is_final,
                chunk_size=chunk_size,
                encoder_chunk_look_back=encoder_chunk_look_back,
                decoder_chunk_look_back=decoder_chunk_look_back,
                )
    res = model(input=speech_chunk, cache=cache, is_final=is_final, chunk_size=chunk_size, encoder_chunk_look_back=encoder_chunk_look_back, decoder_chunk_look_back=decoder_chunk_look_back)
    print(res)
```

@@ -160,11 +154,7 @@
for i in range(total_chunk_num):
    speech_chunk = speech[i*chunk_stride:(i+1)*chunk_stride]
    is_final = i == total_chunk_num - 1
    res = model(input=speech_chunk,
                cache=cache,
                is_final=is_final,
                chunk_size=chunk_size,
                )
    res = model(input=speech_chunk, cache=cache, is_final=is_final, chunk_size=chunk_size)
    if len(res[0]["value"]):
        print(res)
```
@@ -187,8 +177,7 @@

wav_file = f"{model.model_path}/example/asr_example.wav"
text_file = f"{model.model_path}/example/asr_example.wav"
res = model(input=(wav_file, text_file),
            data_type=("sound", "text"))
res = model(input=(wav_file, text_file), data_type=("sound", "text"))
print(res)
```
更多详细用法（[示例](examples/industrial_data_pretraining)）