From 8adcecd19ad2cba330a89d6d9a2744bd97f8846f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E5=88=98=E6=82=A6?= Date: Fri, 26 Jan 2024 15:40:14 +0800 Subject: [PATCH 1/2] =?UTF-8?q?=E5=A2=9E=E5=8A=A0=E5=BC=95=E5=AF=BC?= =?UTF-8?q?=E9=9F=B3=E9=A2=91=E8=AE=B0=E5=BF=86=E3=80=81=E5=BC=95=E5=AF=BC?= =?UTF-8?q?=E9=9F=B3=E9=A2=91=E6=96=87=E6=9C=AC=E8=AE=B0=E5=BF=86=E3=80=81?= =?UTF-8?q?=E8=AF=AD=E7=A7=8D=E8=AE=B0=E5=BF=86?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 增加引导音频记忆、引导音频文本记忆、语种记忆 --- GPT_SoVITS/inference_webui.py | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) diff --git a/GPT_SoVITS/inference_webui.py b/GPT_SoVITS/inference_webui.py index fdee8d9..9a9dd10 100644 --- a/GPT_SoVITS/inference_webui.py +++ b/GPT_SoVITS/inference_webui.py @@ -64,6 +64,23 @@ elif torch.backends.mps.is_available(): else: device = "cpu" +# 操作记忆功能 + +file_path = './audio_log.txt' + +upload_audio_path = None +upload_audio_text = "" +upload_audio_lanuage = "中文" + +if os.path.exists(file_path): + with open(file_path, 'r',encoding="utf-8") as file: + text_data = file.read() + text_data = text_data.split("|") + + upload_audio_path = text_data[0] + upload_audio_text = text_data[1] + upload_audio_lanuage = text_data[2] + tokenizer = AutoTokenizer.from_pretrained(bert_path) bert_model = AutoModelForMaskedLM.from_pretrained(bert_path) if is_half == True: @@ -183,6 +200,7 @@ dict_language={ def get_tts_wav(ref_wav_path, prompt_text, prompt_language, text, text_language): + with open("./audio_log.txt","w",encoding="utf-8")as f:f.write(f"{wav_path_log}|{prompt_text}|{prompt_language}") t0 = ttime() prompt_text = prompt_text.strip("\n") prompt_language, text = prompt_language, text.strip("\n") From 16196b6f38fe1dd381440fabb434976669d47b40 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E5=88=98=E6=82=A6?= Date: Fri, 26 Jan 2024 15:42:49 +0800 Subject: [PATCH 2/2] =?UTF-8?q?=E5=A2=9E=E5=8A=A0=E5=BC=95=E5=AF=BC?= =?UTF-8?q?=E9=9F=B3=E9=A2=91=E8=AE=B0=E5=BF=86=E3=80=81=E5=BC=95=E5=AF=BC?= =?UTF-8?q?=E9=9F=B3=E9=A2=91=E6=96=87=E6=9C=AC=E8=AE=B0=E5=BF=86=E3=80=81?= =?UTF-8?q?=E8=AF=AD=E7=A7=8D=E8=AE=B0=E5=BF=86?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 增加引导音频记忆、引导音频文本记忆、语种记忆 --- GPT_SoVITS/inference_webui.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/GPT_SoVITS/inference_webui.py b/GPT_SoVITS/inference_webui.py index 9a9dd10..3b10fa0 100644 --- a/GPT_SoVITS/inference_webui.py +++ b/GPT_SoVITS/inference_webui.py @@ -418,10 +418,10 @@ with gr.Blocks(title="GPT-SoVITS WebUI") as app: GPT_dropdown.change(change_gpt_weights,[GPT_dropdown],[]) gr.Markdown(value=i18n("*请上传并填写参考信息")) with gr.Row(): - inp_ref = gr.Audio(label=i18n("请上传参考音频"), type="filepath") - prompt_text = gr.Textbox(label=i18n("参考音频的文本"), value="") + inp_ref = gr.Audio(label=i18n("请上传参考音频"), type="filepath",value=upload_audio_path) + prompt_text = gr.Textbox(label=i18n("参考音频的文本"), value=upload_audio_text) prompt_language = gr.Dropdown( - label=i18n("参考音频的语种"),choices=[i18n("中文"),i18n("英文"),i18n("日文")],value=i18n("中文") + label=i18n("参考音频的语种"),choices=[i18n("中文"),i18n("英文"),i18n("日文")],value=i18n(upload_audio_lanuage) ) gr.Markdown(value=i18n("*请填写需要合成的目标文本")) with gr.Row():