Voice repeat after finishing speaking in real-time speaking mode

[gahdeer]

Describe the bug

I downloaded the library and ran it, and it works but there is a problem with the real time speaking mode, where if I speak one word it converts it with the added model voice and keeps repeating it and does not stop even though I am not speaking!

To Reproduce

I use a special mod that I trained myself, everything works as it should, except for speaking in real time, I have a problem with repeating speech.

my config:

Event config_path, values {'model_path': gui.py:665
'C:/Users/m/model/G_3601.pth', 'model_path_browse':
'C:/Users/m/model/G_3601.pth', 'config_path':
'C:/Users/m/model/config.json', 'config_path_browse':
'C:/Users/m/model/config.json', 'cluster_model_path': '',
'cluster_model_path_browse': '', 'speaker': '', 'silence_threshold': -35.0, 'transpose':
12.0, 'auto_predict_f0': True, 'f0_method': 'crepe', 'cluster_infer_ratio': 0.0,
'noise_scale': 0.4, 'pad_seconds': 0.1, 'chunk_seconds': 0.5, 'max_chunk_seconds': 40.0,
'absolute_thresh': True, 'input_path': '', 'input_path_browse': '',
'input_path_folder_browse': '', 'output_path': '', 'output_path_browse': '',
'auto_play': True, 'crossfade_seconds': 0.05, 'block_seconds': 0.35,
'additional_infer_before_seconds': 0.15, 'additional_infer_after_seconds': 0.1,
'realtime_algorithm': '1 (Divide constantly)', 'input_device': 'Microphone (High
Definition Aud (MME)', 'output_device': 'Headphones (High Definition Aud (MME)', 'passthrough_original': False, 'presets':
'Default VC (GPU, GTX 1060)', 'preset_name': '', 'use_gpu': True}

Additional context

python 3.10
so-vits-svc-fork version: 4.1.1

Version

4.1.1

Platform

Windows 10

Code of Conduct

• I agree to follow this project's Code of Conduct.

No Duplicate

• I have checked existing issues to avoid duplicates.