Upload folder using huggingface_hub

11LabsTTS.py

@@ -0,0 +1,168 @@
+"""
+import requests
+import gradio as gr
+
+CHUNK_SIZE = 1024
+
+headers1 = {
+    "Accept": "application/json",
+    "xi-api-key": "54f884da3108725f26af02d5907d1eb4"
+}
+
+headers2 = {
+    "Accept": "audio/mpeg",
+    "Content-Type": "application/json",
+    "xi-api-key": "54f884da3108725f26af02d5907d1eb4"
+}
+
+name_list = []
+
+def elevenlabs_tts(text, voice_name):
+    # 获取声音列表
+    url1 = "https://api.elevenlabs.io/v1/voices"
+    response1 = requests.get(url1, headers=headers1)
+    voices = response1.json()['voices']
+
+    for voice in voices:
+        vid = voice['voice_id']
+        vname = voice['name']
+        label = voice['labels']
+        info = {"voice_id": vid, "name": vname, "labels": label}
+
+        name_list.append(info["name"])
+
+        if info['name'] == voice_name:
+            voice_id = info['voice_id']
+            url2 = "https://api.elevenlabs.io/v1/text-to-speech/" + voice_id
+        # print(infos)
+        # return infos
+
+    # 根据指定的人名合成语音
+    data = {
+        "text": text,
+        "model_id": "eleven_monolingual_v1",
+        "voice_settings": {
+            "stability": 0.5,
+            "similarity_boost": 0.5
+        }
+    }
+
+    response2 = requests.post(url2, json=data, headers=headers2)
+    with open('output.mp3', 'wb') as f:
+        for chunk in response2.iter_content(chunk_size=CHUNK_SIZE):
+            if chunk:
+                f.write(chunk)
+
+    return 'output.mp3'
+
+demo = gr.Interface(
+    fn = elevenlabs_tts,
+
+    # demo输入设置
+    inputs = [
+        gr.Dropdown(name_list, label="发音人"),
+        gr.Textbox(label="输入文本"),
+    ],
+    # demo输出设置
+    outputs = [
+        "audio",
+        "text",
+    ],
+    # demo其他设置
+    title = "Text to Speech Synthesis",
+    description = "Synthesize speech from text using Elevenlabs",
+    examples = [
+        ["Rachel", "Hello World!"],
+        ["Clyde", "This is a test."],
+        ["Domi", "Gradio is awesome!"],
+    ]
+)
+
+if __name__ == "__main__":
+    demo.launch(share=True, server_name='0.0.0.0', server_port=8121)
+    #print(name_list)
+
+"""
+
+import requests
+import gradio as gr
+
+CHUNK_SIZE = 1024
+
+headers1 = {
+    "Accept": "application/json",
+    "xi-api-key": "54f884da3108725f26af02d5907d1eb4"
+}
+
+headers2 = {
+    "Accept": "audio/mpeg",
+    "Content-Type": "application/json",
+    "xi-api-key": "54f884da3108725f26af02d5907d1eb4"
+}
+
+def get_voice_names():
+    url1 = "https://api.elevenlabs.io/v1/voices"
+    response1 = requests.get(url1, headers=headers1)
+    voices = response1.json()['voices']
+    names = [voice['name'] for voice in voices]
+    return names
+
+name_list = get_voice_names()
+
+def elevenlabs_tts(voice_name, text):
+    # 获取声音列表
+    url1 = "https://api.elevenlabs.io/v1/voices"
+    response1 = requests.get(url1, headers=headers1)
+    voices = response1.json()['voices']
+    #print(voice_name)
+    #print(voices)
+
+    for voice in voices:
+        if voice['name'] == voice_name:
+            voice_id = voice['voice_id']
+            label = voice['labels']
+            url2 = "https://api.elevenlabs.io/v1/text-to-speech/" + voice_id
+            #print(voice_id)
+            #print(label)
+            break
+
+    data = {
+        "text": text,
+        "model_id": "eleven_monolingual_v1",
+        "voice_settings": {
+            "stability": 0.5,
+            "similarity_boost": 0.5
+        }
+    }
+
+    response2 = requests.post(url2, json=data, headers=headers2)
+    with open('output.mp3', 'wb') as f:
+        for chunk in response2.iter_content(chunk_size=CHUNK_SIZE):
+            if chunk:
+                f.write(chunk)
+
+    return 'output.mp3', label
+
+demo = gr.Interface(
+    fn = elevenlabs_tts,
+    inputs = [
+        gr.Dropdown(name_list, label="发音人"),
+        gr.Textbox(label="输入文本"),
+    ],
+    outputs = [
+        "audio",
+        "text",
+    ],
+    title = "Text to Speech Synthesis",
+    description = "Synthesize speech from text using Elevenlabs",
+    examples = [
+        ["Rachel", "Hello World!"],
+        ["Clyde", "This is a test."],
+        ["Domi", "Gradio is awesome!"],
+    ]
+)
+
+if __name__ == "__main__":
+    demo.launch(share=True, server_name='0.0.0.0', server_port=8121)
+
+

README.md

@@ -1,12 +1,6 @@
 ---
-title: Gradio
-emoji: 🦀
-colorFrom: blue
-colorTo: purple
+title: gradio
+app_file: gradio_demo.py
 sdk: gradio
 sdk_version: 4.36.1
-app_file: app.py
-pinned: false
 ---
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

api.json

@@ -0,0 +1,12 @@
+{
+  "type": "service_account",
+  "project_id": "gravity-362603",
+  "private_key_id": "aa2ae4651df33c1aab0670e07e265992f943c292",
+  "private_key": "-----BEGIN PRIVATE KEY-----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\nMbI0XZMpHoo60OECydP7Ng==\n-----END PRIVATE KEY-----\n",
+  "client_email": "[email protected]",
+  "client_id": "112160328750047841519",
+  "auth_uri": "https://accounts.google.com/o/oauth2/auth",
+  "token_uri": "https://oauth2.googleapis.com/token",
+  "auth_provider_x509_cert_url": "https://www.googleapis.com/oauth2/v1/certs",
+  "client_x509_cert_url": "https://www.googleapis.com/robot/v1/metadata/x509/gcp-tts-gravity%40gravity-362603.iam.gserviceaccount.com"
+}

ensemble_app.py

@@ -0,0 +1,439 @@
+#!/usr/bin/env python3
+# -*- coding=utf8 -*-
+########################################################################
+#
+# Copyright (c) 2023 Baidu.com, Inc. All Rights Reserved
+#
+########################################################################
+
+"""
+Author: linxiaolong
+"""
+import warnings
+warnings.filterwarnings("ignore")
+
+# 外部库
+import re
+import requests
+import argparse
+import json
+import os
+import re
+import tempfile
+
+import librosa
+import numpy as np
+# import torch
+# from torch import no_grad, LongTensor
+import commons
+import gradio as gr
+import gradio.utils as gr_utils
+import gradio.processing_utils as gr_processing_utils
+
+# 内部库
+from models import SynthesizerTrn
+from text import text_to_sequence, text_to_sequence_for_test, _clean_text
+from mel_processing import spectrogram_torch
+import utils
+from text.symbols import symbols
+
+limitation = os.getenv("SYSTEM") == "spaces"  # limit text and audio length in huggingface spaces
+punct_regex = re.compile(r"[\.!\?。！？]")
+silence_duration = 200
+
+
+def split_text(text, regex):
+    """Split text into sentences by puncutations.
+
+    Args:
+        text: long text.
+        regex: puncutation regex.
+
+    Returns:
+        list of sentences.
+    """
+    sentences = re.split(regex, text)
+    puncts = re.findall(regex, text)
+
+    for i, sentence in enumerate(sentences):
+        if sentence == "":
+            continue
+        if i < len(puncts):
+            sentences[i] = sentences[i] + puncts[i]
+        else:
+            sentences[i] = sentences[i] + "。"
+    sentences = [i for i in sentences if i != ""]
+    return sentences
+
+
+def concat_audio(audio_list, sampling_rate=22050, silence_duration=1000):
+    """Concatenate audio files and insert silence between them.
+
+    Args:
+        audio_list: list of audio files.
+        sampling_rate: audio sampling rate. Defaults to 22050.
+        silence_duration: silence duration in miliseconds. Defaults to 1000.
+
+    Returns:
+        concatenated audio.
+    """
+    silence_samples = int(sampling_rate * silence_duration / 1000)
+    silence = np.zeros(silence_samples, dtype=np.float16)
+
+    audio_num = len(audio_list)
+    if audio_num < 2:
+        return audio_list[0]
+    audio_cat = audio_list[0]
+    for i in range(1, audio_num):
+        audio_cat = np.concatenate((audio_cat, silence, audio_list[i]), axis=0)
+
+    return audio_cat
+
+
+### 外部TTS的超参数
+microsoft_url = "https://japaneast.tts.speech.microsoft.com/cognitiveservices/v1"
+microsoft_headers = {'Content-Type': 'application/json; charset=utf-8', 
+'Ocp-Apim-Subscription-Key':'1f1ef0ce53b84261be94fab81df7e628'}
+microsoft_model_list = [
+    "ja-JP-NanamiNeural", 
+    "ja-JP-KeitaNeural", 
+    "ja-JP-AoiNeural",
+    "ja-JP-DaichiNeural",
+    "ja-JP-MayuNeural",
+    "ja-JP-NaokiNeural",
+    "ja-JP-ShioriNeural"
+]
+
+google_url = "http://gbu.jp02-a30-apisix-sandbox.baidu-int.com/gbu/rest/v2/tts/voice_gq"
+google_headers = {'Content-Type': 'application/json; charset=utf-8', 
+'apikey':'synclub-2383kjhjksxfv.2341gs'}
+google_model_list = [
+    "ja-JP-Neural2-B",
+    "ja-JP-Neural2-C",
+    "ja-JP-Neural2-D",
+    "ja-JP-Standard-A",
+    "ja-JP-Standard-B",
+    "ja-JP-Standard-C",
+    "ja-JP-Standard-D",
+    "ja-JP-Wavenet-A",
+    "ja-JP-Wavenet-B",
+    "ja-JP-Wavenet-C",
+    "ja-JP-Wavenet-D"
+]
+
+coefont_url = "http://gbu.jp02-a30-apisix-sandbox.baidu-int.com/gbu/rest/v2/tts/avatar_coe"
+coefont_headers = {'Content-Type': 'application/json; charset=utf-8', 
+'apikey':'synclub-2383kjhjksxfv.2341gs'}
+coefont_id = [
+    '3f84b7b1-30fb-4677-a704-fd136515303e',
+    '9b826785-bea5-4740-b4cd-e9a286264705',
+    '7632cba3-4aca-4cee-9d15-ad1ac31f670c',
+    '2c91238a-96f9-4cb6-a69a-461ee66b0e6d',
+    '08428dee-65b6-490e-a3a3-60dfcdda889d',
+    'c88367bc-5954-426b-a1ba-a683202803c8',
+    'fb64a764-91d5-4510-bddd-70df3d62709a',
+    '5cfa1f33-bca8-4489-bcbe-701045993162',
+    '94cf7792-7c0c-4be4-88e7-c30d26ab6616',
+    '81dbd387-6ad6-4b22-93f9-4e2a0091b2fe',
+    '931a8568-039a-4cef-add7-bee71629c00e',
+    'f91a9d29-c8b4-443f-ba07-82e7e36bd20b',
+    '23c76cf0-bee0-47fa-b735-9b7bdba9f26a',
+    'cf5fdfb8-85ea-41e1-915b-257936791f17',
+    '0f7b53df-3c24-46a5-84d1-cbea39a956c0',
+    '3d499385-d331-4cbb-93c0-2057e60eddcf',
+    '18ca2f7b-97ca-486d-8f47-858965833642',
+    '33e0a2ff-5050-434c-9506-defe97e52f15',
+    '516b0f32-8b5f-48c5-b60e-38d508e2b06b',
+    'c8720caf-2d2d-4130-8831-92f61f9e25e8',
+    '710001f5-e6f5-4cc0-8ba2-e6aa6da8d807',
+    'd36f8bb1-8bd8-4e90-964a-9dbd3e374093',
+    '2157796c-fe48-4688-b7cc-7ea554edf77d',
+    '5cc0dc91-0c6a-4c50-b7d8-f3117cfe44ef',
+    'be5c5295-aba2-4055-a9da-8926da7fb5a0',
+    '76763239-af14-4c0d-9435-956f096f77dc',
+    '10d298ee-ebbf-4838-a6c5-d608f2e3c338',
+    '694cb06e-73bd-43c4-94d4-f775ad3dbb26',
+    '5cf07e7c-5b1c-4360-a8de-7c928580d4b5',
+    '76e2ba06-b23a-4bbe-8148-e30ede9001b9',
+    'c25ed97f-78f7-4e8f-b2fa-f8e29633588b',
+    'e26382ba-2ae2-4cf7-8c1b-420ab4b845d8',
+    '82c4fcf5-d0ee-4fe9-9b0d-89a65d04f290'
+ ]
+coefont_model_list = [
+    'Canel',
+    '胡麻ちゃん',
+    'バーチャル悪霊',
+    '引寄\u3000法則',
+    'にっし～☆',
+    '志水 智（Tomo Shimizu）',
+    '花撫シア-最高精度-しっかり読み上げ',
+    'UNF/UserNotFound',
+    'RoBaKu',
+    'おにもち',
+    '小菅 将太',
+    '秋月つむぎ（落ち着いたナレーション）',
+    '碧海紘斗_OhmiHiroto',
+    'ちくわぶえ',
+    'unnamed',
+    '今井瑶子(高精度。MC ナレーター 落ち着いたトーンです)',
+    '皆のお母さん',
+    '後藤邑子',
+    '田中和彦',
+    'KTNR',
+    '天渡\u3000早苗',
+    '須戸ゼロ',
+    'とり藻々',
+    '武田 祐子',
+    '【PRO】落ち着きナレーション♯畑耕平',
+    '音暖ののん Ver2.0（最高精度）',
+    'ろさちゃん-soft-v2[最高精度] ¦ Losa-chan -soft- ∀ -汎用式概念χ',
+    'パイナップル秀夫お姉さん',
+    'minamo',
+    'あさのゆき',
+    '聲華 琴音【紡】',
+    '黄琴海月【うるとら】',
+    '高橋 俊輔']
+coefont_id_model_name_dict = dict(zip(coefont_model_list, coefont_id))
+
+all_example = "今日は天気がいいから、一緒にハイキングに行きましょう。"
+
+# def audio_postprocess(self, y):
+#     """
+#     修改gr的音频后处理函数
+#     :param self:
+#     :param y:
+#     :return:
+#     """
+#     if y is None:
+#         return None
+
+#     if gr_utils.validate_url(y):
+#         file = gr_processing_utils.download_to_file(y, dir=self.temp_dir)
+#     elif isinstance(y, tuple):
+#         sample_rate, data = y
+#         file = tempfile.NamedTemporaryFile(
+#             suffix=".wav", dir=self.temp_dir, delete=False
+#         )
+#         gr_processing_utils.audio_to_file(sample_rate, data, file.name)
+#     else:
+#         file = gr_processing_utils.create_tmp_copy_of_file(y, dir=self.temp_dir)
+
+#     return gr_processing_utils.encode_url_or_file_to_base64(file.name)
+
+# gr.Audio.postprocess = audio_postprocess
+
+def get_text(text, hps):
+    """
+    :param text:
+    :param hps:
+    :param is_symbol:
+    :return:
+    """
+    # hps中没有包括symbols
+    text_norm = text_to_sequence(text, hps.data.text_cleaners)
+    # hps中有包括symbols
+    # text_norm = text_to_sequence_for_test(text, hps.symbols, hps.data.text_cleaners)
+    if hps.data.add_blank:
+        text_norm = commons.intersperse(text_norm, 0)
+    text_norm = LongTensor(text_norm)
+    return text_norm
+
+
+def create_tts_fn(model, hps):
+    """
+    :param model:
+    :param hps:
+    :param speaker_ids:
+    :return:
+    """
+    def tts_fn(text, speed, noise_scale=.667, noise_scale_w=0.8, volume=1.0):
+        """
+        :param text:
+        :param speaker:
+        :param speed:
+        :param emo:
+        :param volume:
+        :param is_symbol:
+        :return:
+        """
+        sentences = split_text(text, punct_regex)
+        audio_list = []
+        for sentence in sentences:
+            stn_tst = get_text(sentence, hps)
+            with no_grad():
+                x_tst = stn_tst.unsqueeze(0).to(device)
+                x_tst_lengths = LongTensor([stn_tst.size(0)]).to(device)
+                audio = model.infer(x_tst, x_tst_lengths, noise_scale=noise_scale, noise_scale_w=noise_scale_w,
+                                    length_scale=1.0 / speed)[0][0, 0].data.cpu().float().numpy()
+                audio_list.append(audio)
+            del stn_tst, x_tst, x_tst_lengths
+        audio = concat_audio(audio_list, hps.data.sampling_rate, silence_duration)
+        audio = audio * volume
+        return "Success", (hps.data.sampling_rate, audio)
+    return tts_fn
+
+
+def microsoft(text, name, style="Neural"):
+    """
+    :param text:
+    :param name:
+    :param style:
+    :return:
+    """
+    headers = {
+        'Ocp-Apim-Subscription-Key': '1f1ef0ce53b84261be94fab81df7e628',
+        'Content-Type': 'application/ssml+xml',
+        'X-Microsoft-OutputFormat': 'audio-16khz-128kbitrate-mono-mp3',
+        'User-Agent': 'curl',
+    }
+
+    data = ("<speak version='1.0' xml:lang='en-US'>"
+        f"<voice xml:lang='en-US' name='{name}'>" # xml:gender='Female' 
+        f"{text}"
+        "</voice>"
+        "</speak>")
+
+    response = requests.post(
+        'https://japaneast.tts.speech.microsoft.com/cognitiveservices/v1',
+        headers=headers,
+        data=data,
+        proxies= {
+            'http': 'http://192.168.3.11:80',
+            'https': 'http://192.168.3.11:80',
+        }
+    )
+    data = {
+        "text":text,
+        "name":name,
+        "style":style,
+        "format":"mp3"}
+    audio_url = requests.get(microsoft_url, headers=microsoft_headers, json=data).json()['data']['url']
+    return "Success", audio_url
+    
+
+def google(text, name):
+    """
+    :param text:
+    :param name:
+    :param style:
+    :return:
+    """
+    data = {
+        "text":text,
+        "name":name,
+        "sample_rate":16000}
+    audio_url = requests.get(google_url, headers=google_headers, json=data).json()['data']['url']
+    return "Success", audio_url
+
+
+def coefont(text, name):
+    """
+    :param text:
+    :param name:
+    :param style:
+    :return:
+    """
+    data = {
+        "text":text,
+        "coefont":coefont_id_model_name_dict[name]
+        }
+    audio_url = requests.get(coefont_url, headers=coefont_headers, json=data).json()['data']['url']
+    return "Success", audio_url
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--device', type=str, default='cuda')
+    parser.add_argument("--share", action="store_true", default=False, help="share gradio app")
+    parser.add_argument("--port", type=int, default=8080, help="port")
+    parser.add_argument('--model_info_path', type=str, default='/gluster/speech_data/info.json')
+    args = parser.parse_args()
+
+    device = torch.device(args.device)
+    models_tts = []
+
+    with open(args.model_info_path, "r", encoding="utf-8") as f:
+        models_info = json.load(f)
+    for i, info in models_info.items():
+        model_name = info["model_name"]
+        author = info["author"]
+        lang = info["lang"]
+        example = info["example"]
+        config_path = info["config_path"]
+        model_path = info["model_path"]
+        model_type = info["model_type"]
+
+        hps = utils.get_hparams_from_file(config_path)
+        if model_type == "vits":
+            emotion_type = None
+        elif model_type == "vits-emotion":
+            emotion_type = "embedding"
+        elif model_type == "vits-emotion-logits":
+            emotion_type = "logits"
+
+        model = SynthesizerTrn(
+            len(symbols),
+            hps.data.filter_length // 2 + 1,
+            hps.train.segment_size // hps.data.hop_length,
+            emotion_type=emotion_type,
+            **hps.model)
+
+        utils.load_checkpoint(model_path, model, None)
+        model.eval().to(device)
+        if model_type == "vits":
+            # 普通TTS
+            models_tts.append((model_name, author, lang, example, create_tts_fn(model, hps)))
+    app = gr.Blocks()
+    with app:
+        gr.Markdown("## Japanese TTS Demo")
+        with gr.Tabs():
+            with gr.TabItem("自研"):
+                with gr.Tabs():
+                    for i, (model_name, author, lang, example, tts_fn) in enumerate(models_tts):
+                        with gr.TabItem(model_name):
+                            with gr.Column():
+                                tts_input1 = gr.TextArea(label="Text", value=example)
+                                tts_input2 = gr.Slider(label="Speed", value=1.0, minimum=0.4, maximum=3, step=0.1)
+                                tts_input3 = gr.Slider(label="noise_scale", value=0.0, minimum=0.0, maximum=2, step=0.1)
+                                tts_input4 = gr.Slider(label="noise_scale_w", value=0.0,
+                                                       minimum=0.0, maximum=2, step=0.1)
+                                tts_input5 = gr.Slider(label="volume", value=1.0, minimum=0.1, maximum=4, step=0.1)                                
+                                tts_submit = gr.Button("Generate", variant="primary")
+                                tts_output1 = gr.Textbox(label="Output Message")
+                                tts_output2 = gr.Audio(label="Output Audio")
+                                tts_submit.click(tts_fn, [tts_input1, tts_input2, tts_input3, tts_input4, tts_input5],
+                                                 [tts_output1, tts_output2])
+            
+            with gr.TabItem("谷歌"):
+                tts_input1 = gr.TextArea(label="Text", value=all_example)
+                tts_input2 = gr.Dropdown(google_model_list, label="name")                             
+                tts_submit = gr.Button("Generate", variant="primary")
+                tts_output1 = gr.Textbox(label="Output Message")
+                tts_output2 = gr.Audio(label="Output Audio")
+                tts_submit.click(google, [tts_input1, tts_input2],
+                                    [tts_output1, tts_output2])
+
+            with gr.TabItem("微软"):
+                tts_input1 = gr.TextArea(label="Text", value=all_example)
+                tts_input2 = gr.Dropdown(microsoft_model_list, label="name")                             
+                tts_submit = gr.Button("Generate", variant="primary")
+                tts_output1 = gr.Textbox(label="Output Message")
+                tts_output2 = gr.Audio(label="Output Audio")
+                tts_submit.click(microsoft, [tts_input1, tts_input2],
+                                    [tts_output1, tts_output2])
+                
+            with gr.TabItem("coefont"):
+                tts_input1 = gr.TextArea(label="Text", value=all_example)
+                tts_input2 = gr.Dropdown(coefont_model_list, label="name")
+                tts_submit = gr.Button("Generate", variant="primary")
+                tts_output1 = gr.Textbox(label="Output Message")
+                tts_output2 = gr.Audio(label="Output Audio")
+                tts_submit.click(coefont, [tts_input1, tts_input2],
+                                    [tts_output1, tts_output2])
+
+    app.queue(concurrency_count=5).launch(show_api=False,
+                                          share=args.share,
+                                          server_name='0.0.0.0',
+                                          server_port=args.port,
+                                          show_error=True)
+

gradio_demo.py

@@ -0,0 +1,13 @@
+import gradio as gr
+def greet(name):
+    return "Hello " + name + "!"
+with gr.Blocks() as demo:
+    # 设置输入组件
+    name = gr.Textbox(label="Name")
+    # 设置输出组件
+    output = gr.Textbox(label="Output Box")
+    # 设置按钮
+    greet_btn = gr.Button("Greet")
+    # 设置按钮点击事件，默认是垂直
+    greet_btn.click(fn=greet, inputs=name, outputs=output)
+demo.launch(share=True)

guge_api.py

@@ -0,0 +1,32 @@
+import requests
+import base64
+import json
+import os
+os.environ["GOOGLE_APPLICATION_CREDENTIALS"]="./api.json"
+
+url = "https://texttospeech.googleapis.com/v1/text:synthesize"
+headers = {'Content-Type': 'application/json; charset=utf-8','Authorization': 'Bearer ya29.c.b0Aaekm1Iv0qkoIurI3LgzvNRqWE9tZRNusy9HCLT_xcpTXvdNsvYQt_9PUIXDQAvHV-b9abi-_n9Totai8fgJ7qkJ5sJszhU8bKTkP9zPJccfTkU4pCvyJxCVO1APIj7p3Hdvl9LWAIGb6aqFdz0a2dSn2mzsGKSsv0UqjlPr4M4s08HWkW8fqdE8UO98xphUTm0bEYmYFn-feWAPG5QToNi62c_4KZxveCCPhZZOtWLSPCGOj-D9z1YIHZZcdWz8mWQ5lO2WpASEroak4ohRchjSEpnJka8F3uwoY36Oms4M32d9uKYbsdNMFfS1CudzvDYXKSBkH337KZqb6o-zhFUm13Ivxlr1Zbtvv6dw0kbg_my1-p69v9tbXIsgxvgYfjMfhxgxfOX3aJxsgnU-jw0d1w5jgs70InjbZRagcrRth24Sm4wzyma40j5UodiIvI0sRxQOBUzaZ93YXZje1S_vtoxcSUl9VSrQY9rWt2uk2Ili-aIsI6rUa06McJcnSB-jQhq61z3xpoqRXVd04SMZlc1YnWxhZVYakZoRIivMwFkjRvb0SIn7x_1M880QF4IYzMOb7eVkJW2t41u5y2rWSc7rerjetWd2vRehn63vJRp0jn3lskyM5YZqqiWs4VYjqdVt7f9jy3qR2Vgn5V2hvSdrjdbkSgc1UB3fl4k-4sbjrd-M3OpO99R0I921-nMc8lBgIFum9qZ-VcWJJ-ecYZtygSXiiR6ljX6p70h6m90vVbc8bwnk7ez99mUOzhkuR8B8wb1aJd8XyzdYuqwWfMyq7U89dUju-m6dji-XeMuSSwSiueaVhsU68FXiJa_-ieZra1Saak7zjq_feg8hjZRujbh0e_YqRO4Bhf7-5F-UrO1n7XwJIzofR_uxJfgxs6MuFxfUJX87azO1I31Zty5ZSc2Q4mynq0mlQWOvhl7fr8pSJMzfQtOJW6wOc0RrW3ouIB-mWRrie46gbqF0FzSylkipZ3JasoIeO-gS_olF-YfYR5i'
+}
+text = "二月の下旬に差し掛かる頃だった。"
+
+data = {
+    "input":{
+        "text":text
+    },
+    "voice":{
+        "languageCode":"ja-JP",
+        "name":"ja-JP-Neural2-C",
+        "ssmlGender":"MALE"
+    },
+    "audioConfig":{
+        "audioEncoding":"MP3"
+    }
+    }
+response = requests.post(url, headers=headers, json=data)
+print(response)
+# response = response.json()
+audio = response.content
+#audio = base64.b64decode(audio)
+# The response's audio_content is binary.
+with open("test9.mp3", "wb") as out:
+    out.write(audio)

install.sh

@@ -0,0 +1,17 @@
+#!/bin/bash
+
+URL=https://dl.google.com/dl/cloudsdk/channels/rapid/install_google_cloud_sdk.bash
+
+function download {
+  scratch="$(mktemp -d -t tmp.XXXXXXXXXX)" || exit
+  script_file="$scratch/install_google_cloud_sdk.bash"
+
+  echo "Downloading Google Cloud SDK install script: $URL"
+  curl -# "$URL" > "$script_file" || exit
+  chmod 775 "$script_file"
+
+  echo "Running install script from: $script_file"
+  "$script_file" "$@"
+}
+
+download "$@"

microsoftTTS.py

@@ -0,0 +1,31 @@
+import os
+import azure.cognitiveservices.speech as speechsdk
+
+speech_key = "1f1ef0ce53b84261be94fab81df7e628"
+service_region = "japaneast"
+speech_config = speechsdk.SpeechConfig(subscription=speech_key, region=service_region)
+
+# This example requires environment variables named "SPEECH_KEY" and "SPEECH_REGION"
+#speech_config = speechsdk.SpeechConfig(subscription=os.environ.get('SPEECH_KEY'), region=os.environ.get('SPEECH_REGION'))
+audio_config = speechsdk.audio.AudioOutputConfig(use_default_speaker=True)
+
+# The neural multilingual voice can speak different languages based on the input text.
+speech_config.speech_synthesis_voice_name='en-US-AvaMultilingualNeural'
+
+speech_synthesizer = speechsdk.SpeechSynthesizer(speech_config=speech_config, audio_config=audio_config)
+
+# Get text from the console and synthesize to the default speaker.
+print("Enter some text that you want to speak >")
+text = input()
+
+speech_synthesis_result = speech_synthesizer.speak_text_async(text).get()
+
+if speech_synthesis_result.reason == speechsdk.ResultReason.SynthesizingAudioCompleted:
+    print("Speech synthesized for text [{}]".format(text))
+elif speech_synthesis_result.reason == speechsdk.ResultReason.Canceled:
+    cancellation_details = speech_synthesis_result.cancellation_details
+    print("Speech synthesis canceled: {}".format(cancellation_details.reason))
+    if cancellation_details.reason == speechsdk.CancellationReason.Error:
+        if cancellation_details.error_details:
+            print("Error details: {}".format(cancellation_details.error_details))
+            print("Did you set the speech resource key and region values?")

test.py

@@ -0,0 +1,357 @@
+import warnings
+warnings.filterwarnings("ignore")
+
+# 外部库
+import re
+import requests
+import argparse
+import json
+import os
+import re
+import tempfile
+import librosa
+import numpy as np
+# import torch
+# from torch import no_grad, LongTensor
+# import commons
+import gradio as gr
+import gradio.utils as gr_utils
+import gradio.processing_utils as gr_processing_utils
+
+# 内部库
+# from models import SynthesizerTrn
+# from text import text_to_sequence, text_to_sequence_for_test, _clean_text
+# from mel_processing import spectrogram_torch
+# import utils
+# from text.symbols import symbols
+all_example = "my voice is my passport verify me."
+
+eleven_voice_id = [
+    "21m00Tcm4TlvDq8ikWAM",
+    "29vD33N1CtxCmqQRPOHJ",
+    "2EiwWnXFnvU5JabPnv8n",
+    "5Q0t7uMcjvnagumLfvZi",
+    "AZnzlk1XvdvUeBnXmlld",
+    "CYw3kZ02Hs0563khs1Fj",
+    "D38z5RcWu1voky8WS1ja",
+    "EXAVITQu4vr4xnSDxMaL",
+    "ErXwobaYiN019PkySvjV",
+    "GBv7mTt0atIp3Br8iCZE",
+    "IKne3meq5aSn9XLyUdCD",
+    "JBFqnCBsd6RMkjVDRZzb",
+    "LcfcDJNUP1GQjkzn1xUU",
+    "MF3mGyEYCl7XYWbV9V6O",
+    "N2lVS1w4EtoT3dr4eOWO",
+    "ODq5zmih8GrVes37Dizd",
+    "SOYHLrjzK2X1ezoPC6cr",
+    "TX3LPaxmHKxFdv7VOQHJ",
+    "ThT5KcBeYPX3keUQqHPh",
+    "TxGEqnHWrfWFTfGW9XjX",
+    "VR6AewLTigWG4xSOukaG",
+    "XB0fDUnXU5powFXDhCwa",
+    "Xb7hH8MSUJpSbSDYk0k2",
+    "XrExE9yKIg1WjnnlVkGX",
+    "ZQe5CZNOzWyzPSCn5a3c",
+    "Zlb1dXrM653N07WRdFW3",
+    "bVMeCyTHy58xNoL34h3p",
+    "flq6f7yk4E4fJM5XTYuZ",
+    "g5CIjZEefAph4nQFvHAz",
+    "iP95p4xoKVk53GoZ742B",
+    "jBpfuIE2acCO8z3wKNLl",
+    "jsCqWAovK2LkecY7zXl4",
+    "nPczCjzI2devNBz1zQrb",
+    "oWAxZDx7w5VEj9dCyTzz",
+    "onwK4e9ZLuTAKqWW03F9",
+    "pFZP5JQG7iQjIQuC4Bku",
+    "pMsXgVXv3BLzUgSXRplE",
+    "pNInz6obpgDQGcFmaJgB",
+    "piTKgcLEGmPE4e6mEKli",
+    "pqHfZKP75CvOlQylNhV4",
+    "t0jbNlBVZ17f02VDIeMI",
+    "yoZ06aMxZJJ28mfd3POQ",
+    "z9fAnlkpzviPz146aGWa",
+    "zcAOhNBS3c14rBihAFp1",
+    "zrHiDhphv9ZnVXBqCLjz",
+]
+
+eleven_name = [
+    "Rachel",
+    "Drew",
+    "Clyde",
+    "Paul",
+    "Domi",
+    "Dave",
+    "Fin",
+    "Sarah",
+    "Antoni",
+    "Thomas",
+    "Charlie",
+    "George",
+    "Emily",
+    "Elli",
+    "Callum",
+    "Patrick",
+    "Harry",
+    "Liam",
+    "Dorothy",
+    "Josh",
+    "Arnold",
+    "Charlotte",
+    "Alice",
+    "Matilda",
+    "James",
+    "Joseph",
+    "Jeremy",
+    "Michael",
+    "Ethan",
+    "Chris",
+    "Gigi",
+    "Freya",
+    "Brian",
+    "Grace",
+    "Daniel",
+    "Lily",
+    "Serena",
+    "Adam",
+    "Nicole",
+    "Bill",
+    "Jessie",
+    "Sam",
+    "Glinda",
+    "Giovanni",
+    "Mimi",
+]
+eleven_id_model_name_dict = dict(zip(eleven_name, eleven_voice_id))
+
+def openai(text, name):
+
+    headers = {
+        'Authorization': 'Bearer ' + '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',
+        'Content-Type': 'application/json',
+    }
+
+    json_data = {
+        'model': 'tts-1-hd',
+        'input': f'{text}',
+        'voice': f'{name}',
+    }
+
+    response = requests.post('https://api.openai.com/v1/audio/speech', headers=headers, json=json_data)
+
+    # Note: json_data will not be serialized by requests
+    # exactly as it was in the original request.
+    #data = '{\n    "model": "tts-1",\n    "input": "The quick brown fox jumped over the lazy dog.",\n    "voice": "alloy"\n  }'
+    #response = requests.post('https://api.openai.com/v1/audio/speech', headers=headers, data=data)
+
+    return "Success", response
+
+def elevenlabs(text,name):
+    url = f"https://api.elevenlabs.io/v1/text-to-speech/{name}"
+    CHUNK_SIZE = 1024
+    #url = "https://api.elevenlabs.io/v1/text-to-speech/<voice-id>"
+
+    headers = {
+    "Accept": "audio/mpeg",
+    "Content-Type": "application/json",
+    "xi-api-key": "a3391f0e3ff8472b61978dbb70ccc6fe"
+    }
+
+    data = {
+    "text": f"{text}",
+    "model_id": "eleven_monolingual_v1",
+    "voice_settings": {
+        "stability": 0.5,
+        "similarity_boost": 0.5
+    }
+    }
+
+    response = requests.post(url, json=data, headers=headers)
+    # with open('output.mp3', 'wb') as f:
+    #     for chunk in response.iter_content(chunk_size=CHUNK_SIZE):
+    #         if chunk:
+    #             f.write(chunk)
+    return "Success", response
+
+microsoft_model_list = [
+    "en-US-AvaMultilingualNeural"
+]
+
+def microsoft(text, name, style="Neural"):
+    """
+    :param text:
+    :param name:
+    :param style:
+    :return:
+    """
+    headers = {
+        'Ocp-Apim-Subscription-Key': '1f1ef0ce53b84261be94fab81df7e628',
+        'Content-Type': 'application/ssml+xml',
+        'X-Microsoft-OutputFormat': 'audio-16khz-128kbitrate-mono-mp3',
+        'User-Agent': 'curl',
+    }
+
+    data = ("<speak version='1.0' xml:lang='en-US'>"
+        f"<voice xml:lang='en-US' name='{name}'>" # xml:gender='Female' 
+        f"{text}"
+        "</voice>"
+        "</speak>")
+
+    response = requests.post(
+        'https://japaneast.tts.speech.microsoft.com/cognitiveservices/v1',
+        headers=headers,
+        data=data,
+    )
+    # data = {
+    #     "text":text,
+    #     "name":name,
+    #     "style":style,
+    #     "format":"mp3"}
+    # audio_url = requests.get(microsoft_url, headers=microsoft_headers, json=data).json()['data']['url']
+    return "Success", response
+    
+# def google(text,name):
+#     # import subprocess
+#     # command1 = subprocess.run('gcloud auth print-access-token', shell=True, capture_output=True, text=True).stdout
+
+#     headers = {
+#         'Authorization': 'Bearer ' + "synclub-2383kjhjksxfv.2341gs",
+#         'x-goog-user-project': 'PROJECT_ID',
+#         'Content-Type': 'application/json; charset=utf-8',
+#     }
+
+#     data = {
+#             "input": {
+#                 "text": f"{text}"},
+    #         "voice": {
+    #             "languageCode": "en-gb",
+    #             "name": "en-GB-Standard-A",
+    #             "ssmlGender": "FEMALE"
+    #         },
+    #         "audioConfig": {
+    #             "audioEncoding": "MP3"
+    #         }
+    # }
+
+    # response = requests.post('https://texttospeech.googleapis.com/v1/text:synthesize', headers=headers, data=data)
+    # return "Success", response
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--device', type=str, default='cuda')
+    parser.add_argument("--share", action="store_true", default=True, help="share gradio app")
+    parser.add_argument("--port", type=int, default=8081, help="port")
+    parser.add_argument('--model_info_path', type=str, default='/gluster/speech_data/info.json')
+    args = parser.parse_args()
+    # app = gr.Blocks()
+    # with app:
+    #     gr.Markdown("## Japanese TTS Demo")
+    #     with gr.Tabs():
+    #         with gr.TabItem("微软"):
+    #             tts_input1 = gr.TextArea(label="Text", value=all_example)
+    #             tts_input2 = gr.Dropdown(microsoft_model_list, label="name")                             
+    #             tts_submit = gr.Button("Generate", variant="primary")
+    #             tts_output1 = gr.Textbox(label="Output Message")
+    #             tts_output2 = gr.Audio(label="Output Audio")
+    #             tts_submit.click(microsoft, [tts_input1, tts_input2],
+    #                                 [tts_output1, tts_output2])
+    # _, audio = microsoft(all_example, 'en-US-AvaMultilingualNeural')
+    # _, audio = google(all_example,'alloy')
+    # print(audio)
+    # with open("test4.mp3", "wb") as f:
+    #     f.write(audio.content)
+    #_, audio = elevenlabs(all_example, "21m00Tcm4TlvDq8ikWAM")
+    # print(audio)
+    # with open('output.mp3', 'wb') as f:
+    #     for chunk in audio.iter_content(chunk_size=1024):
+    #         if chunk:
+    #             f.write(chunk)
+
+    # device = torch.device(args.device)
+    # models_tts = []
+
+    # with open(args.model_info_path, "r", encoding="utf-8") as f:
+    #     models_info = json.load(f)
+    # for i, info in models_info.items():
+        # model_name = info["model_name"]
+        # author = info["author"]
+        # lang = info["lang"]
+        # example = info["example"]
+        # config_path = info["config_path"]
+        # model_path = info["model_path"]
+        # model_type = info["model_type"]
+
+        # hps = utils.get_hparams_from_file(config_path)
+        # if model_type == "vits":
+        #     emotion_type = None
+        # elif model_type == "vits-emotion":
+        #     emotion_type = "embedding"
+        # elif model_type == "vits-emotion-logits":
+        #     emotion_type = "logits"
+
+        # model = SynthesizerTrn(
+        #     len(symbols),
+        #     hps.data.filter_length // 2 + 1,
+        #     hps.train.segment_size // hps.data.hop_length,
+        #     emotion_type=emotion_type,
+        #     **hps.model)
+
+        # utils.load_checkpoint(model_path, model, None)
+        # model.eval().to(device)
+        # if model_type == "vits":
+        #     # 普通TTS
+        #     models_tts.append((model_name, author, lang, example, create_tts_fn(model, hps)))
+    
+    
+    app = gr.Blocks()
+    with app:
+        gr.Markdown("## Japanese TTS Demo")
+        with gr.Tabs():
+            # with gr.TabItem("自研"):
+            #     with gr.Tabs():
+            #         for i, (model_name, author, lang, example, tts_fn) in enumerate(models_tts):
+            #             with gr.TabItem(model_name):
+            #                 with gr.Column():
+            #                     tts_input1 = gr.TextArea(label="Text", value=example)
+            #                     tts_input2 = gr.Slider(label="Speed", value=1.0, minimum=0.4, maximum=3, step=0.1)
+            #                     tts_input3 = gr.Slider(label="noise_scale", value=0.0, minimum=0.0, maximum=2, step=0.1)
+            #                     tts_input4 = gr.Slider(label="noise_scale_w", value=0.0,
+            #                                            minimum=0.0, maximum=2, step=0.1)
+            #                     tts_input5 = gr.Slider(label="volume", value=1.0, minimum=0.1, maximum=4, step=0.1)                                
+            #                     tts_submit = gr.Button("Generate", variant="primary")
+            #                     tts_output1 = gr.Textbox(label="Output Message")
+            #                     tts_output2 = gr.Audio(label="Output Audio")
+            #                     tts_submit.click(tts_fn, [tts_input1, tts_input2, tts_input3, tts_input4, tts_input5],
+            #                                      [tts_output1, tts_output2])
+            
+            # with gr.TabItem("谷歌"):
+            #     tts_input1 = gr.TextArea(label="Text", value=all_example)
+            #     tts_input2 = gr.Dropdown(google_model_list, label="name")                             
+            #     tts_submit = gr.Button("Generate", variant="primary")
+            #     tts_output1 = gr.Textbox(label="Output Message")
+            #     tts_output2 = gr.Audio(label="Output Audio")
+            #     tts_submit.click(google, [tts_input1, tts_input2],
+            #                         [tts_output1, tts_output2])
+
+            with gr.TabItem("微软"):
+                tts_input1 = gr.TextArea(label="Text", value=all_example)
+                tts_input2 = gr.Dropdown(microsoft_model_list, label="name")                             
+                tts_submit = gr.Button("Generate", variant="primary")
+                tts_output1 = gr.Textbox(label="Output Message")
+                tts_output2 = gr.Audio(label="Output Audio")
+                tts_submit.click(microsoft, [tts_input1, tts_input2],
+                                    [tts_output1, tts_output2])
+                
+            # with gr.TabItem("coefont"):
+            #     tts_input1 = gr.TextArea(label="Text", value=all_example)
+            #     tts_input2 = gr.Dropdown(coefont_model_list, label="name")
+            #     tts_submit = gr.Button("Generate", variant="primary")
+            #     tts_output1 = gr.Textbox(label="Output Message")
+            #     tts_output2 = gr.Audio(label="Output Audio")
+            #     tts_submit.click(coefont, [tts_input1, tts_input2],
+            #                         [tts_output1, tts_output2])
+
+    app.launch(show_api=False,
+                share=args.share,
+                server_name='0.0.0.0',
+                server_port=args.port,
+                show_error=True)
+

test2.mp3

@@ -0,0 +1 @@
+{"detail":{"status":"invalid_uid","message":"An invalid ID has been received: '<string>'. Make sure to provide a correct one."}}

test4.mp3

@@ -0,0 +1,17 @@
+{
+  "error": {
+    "code": 401,
+    "message": "Request had invalid authentication credentials. Expected OAuth 2 access token, login cookie or other valid authentication credential. See https://developers.google.com/identity/sign-in/web/devconsole-project.",
+    "status": "UNAUTHENTICATED",
+    "details": [
+      {
+        "@type": "type.googleapis.com/google.rpc.ErrorInfo",
+        "reason": "ACCESS_TOKEN_TYPE_UNSUPPORTED",
+        "metadata": {
+          "method": "google.cloud.texttospeech.v1.TextToSpeech.SynthesizeSpeech",
+          "service": "texttospeech.googleapis.com"
+        }
+      }
+    ]
+  }
+}

test9.mp3

@@ -0,0 +1,17 @@
+{
+  "error": {
+    "code": 401,
+    "message": "Request had invalid authentication credentials. Expected OAuth 2 access token, login cookie or other valid authentication credential. See https://developers.google.com/identity/sign-in/web/devconsole-project.",
+    "status": "UNAUTHENTICATED",
+    "details": [
+      {
+        "@type": "type.googleapis.com/google.rpc.ErrorInfo",
+        "reason": "ACCESS_TOKEN_TYPE_UNSUPPORTED",
+        "metadata": {
+          "service": "texttospeech.googleapis.com",
+          "method": "google.cloud.texttospeech.v1.TextToSpeech.SynthesizeSpeech"
+        }
+      }
+    ]
+  }
+}

try.py

@@ -0,0 +1,33 @@
+import requests
+import os
+import base64
+import json
+os.environ["GOOGLE_APPLICATION_CREDENTIALS"]="./api.json"
+
+url = "https://texttospeech.googleapis.com/v1/text:synthesize"
+headers = {'Content-Type': 'application/json; charset=utf-8',
+           'X-Goog-Api-Key': 'synclub-2383kjhjksxfv.2341gs' # 待补充
+          }
+text = "二月の下旬に差し掛かる頃だった。"
+
+data = {
+    "input":{
+        "text":text
+    },
+    "voice":{
+        "languageCode":"ja-JP",
+        "name":"ja-JP-Neural2-C",
+        "ssmlGender":"MALE"
+    },
+    "audioConfig":{
+        "audioEncoding":"MP3"
+    }
+    }
+response = requests.post(url, headers=headers, json=data)
+response = response.json()
+print(response)
+audio = response['audioContent']
+audio = base64.b64decode(audio)
+# The response's audio_content is binary.
+with open("test.mp3", "wb") as out:
+    out.write(audio)

tts_gradio.py

@@ -0,0 +1,255 @@
+import warnings
+warnings.filterwarnings("ignore")
+
+# 外部库
+import re
+import requests
+import argparse
+import json
+import os
+import re
+import tempfile
+import librosa
+import numpy as np
+# import torch
+# from torch import no_grad, LongTensor
+# import commons
+import gradio as gr
+import gradio.utils as gr_utils
+import gradio.processing_utils as gr_processing_utils
+
+all_example = "my voice is my passport verify me."
+
+microsoft_model_list = [
+    "en-US-AvaMultilingualNeural"
+]
+
+openai_model_list = [
+    "alloy",
+    "echo",
+    "fable",
+    "onyx",
+    "nova",
+    "shimmer"
+]
+
+eleven_voice_id = [
+    "21m00Tcm4TlvDq8ikWAM",
+    "29vD33N1CtxCmqQRPOHJ",
+    "2EiwWnXFnvU5JabPnv8n",
+    "5Q0t7uMcjvnagumLfvZi",
+    "AZnzlk1XvdvUeBnXmlld",
+    "CYw3kZ02Hs0563khs1Fj",
+    "D38z5RcWu1voky8WS1ja",
+    "EXAVITQu4vr4xnSDxMaL",
+    "ErXwobaYiN019PkySvjV",
+    "GBv7mTt0atIp3Br8iCZE",
+    "IKne3meq5aSn9XLyUdCD",
+    "JBFqnCBsd6RMkjVDRZzb",
+    "LcfcDJNUP1GQjkzn1xUU",
+    "MF3mGyEYCl7XYWbV9V6O",
+    "N2lVS1w4EtoT3dr4eOWO",
+    "ODq5zmih8GrVes37Dizd",
+    "SOYHLrjzK2X1ezoPC6cr",
+    "TX3LPaxmHKxFdv7VOQHJ",
+    "ThT5KcBeYPX3keUQqHPh",
+    "TxGEqnHWrfWFTfGW9XjX",
+    "VR6AewLTigWG4xSOukaG",
+    "XB0fDUnXU5powFXDhCwa",
+    "Xb7hH8MSUJpSbSDYk0k2",
+    "XrExE9yKIg1WjnnlVkGX",
+    "ZQe5CZNOzWyzPSCn5a3c",
+    "Zlb1dXrM653N07WRdFW3",
+    "bVMeCyTHy58xNoL34h3p",
+    "flq6f7yk4E4fJM5XTYuZ",
+    "g5CIjZEefAph4nQFvHAz",
+    "iP95p4xoKVk53GoZ742B",
+    "jBpfuIE2acCO8z3wKNLl",
+    "jsCqWAovK2LkecY7zXl4",
+    "nPczCjzI2devNBz1zQrb",
+    "oWAxZDx7w5VEj9dCyTzz",
+    "onwK4e9ZLuTAKqWW03F9",
+    "pFZP5JQG7iQjIQuC4Bku",
+    "pMsXgVXv3BLzUgSXRplE",
+    "pNInz6obpgDQGcFmaJgB",
+    "piTKgcLEGmPE4e6mEKli",
+    "pqHfZKP75CvOlQylNhV4",
+    "t0jbNlBVZ17f02VDIeMI",
+    "yoZ06aMxZJJ28mfd3POQ",
+    "z9fAnlkpzviPz146aGWa",
+    "zcAOhNBS3c14rBihAFp1",
+    "zrHiDhphv9ZnVXBqCLjz",
+]
+
+eleven_name = [
+    "Rachel",
+    "Drew",
+    "Clyde",
+    "Paul",
+    "Domi",
+    "Dave",
+    "Fin",
+    "Sarah",
+    "Antoni",
+    "Thomas",
+    "Charlie",
+    "George",
+    "Emily",
+    "Elli",
+    "Callum",
+    "Patrick",
+    "Harry",
+    "Liam",
+    "Dorothy",
+    "Josh",
+    "Arnold",
+    "Charlotte",
+    "Alice",
+    "Matilda",
+    "James",
+    "Joseph",
+    "Jeremy",
+    "Michael",
+    "Ethan",
+    "Chris",
+    "Gigi",
+    "Freya",
+    "Brian",
+    "Grace",
+    "Daniel",
+    "Lily",
+    "Serena",
+    "Adam",
+    "Nicole",
+    "Bill",
+    "Jessie",
+    "Sam",
+    "Glinda",
+    "Giovanni",
+    "Mimi",
+]
+eleven_id_model_name_dict = dict(zip(eleven_name, eleven_voice_id))
+
+def openai(text, name):
+
+    headers = {
+        'Authorization': 'Bearer ' + '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',
+        'Content-Type': 'application/json',
+    }
+
+    json_data = {
+        'model': 'tts-1-hd',
+        'input': f'{text}',
+        'voice': f'{name}',
+    }
+
+    response = requests.post('https://api.openai.com/v1/audio/speech', headers=headers, json=json_data)
+
+    # Note: json_data will not be serialized by requests
+    # exactly as it was in the original request.
+    #data = '{\n    "model": "tts-1",\n    "input": "The quick brown fox jumped over the lazy dog.",\n    "voice": "alloy"\n  }'
+    #response = requests.post('https://api.openai.com/v1/audio/speech', headers=headers, data=data)
+
+    return "Success", response
+
+def elevenlabs(text,name):
+    url = f"https://api.elevenlabs.io/v1/text-to-speech/{eleven_id_model_name_dict[name]}"
+    CHUNK_SIZE = 1024
+    #url = "https://api.elevenlabs.io/v1/text-to-speech/<voice-id>"
+
+    headers = {
+    "Accept": "audio/mpeg",
+    "Content-Type": "application/json",
+    "xi-api-key": "a3391f0e3ff8472b61978dbb70ccc6fe"
+    }
+
+    data = {
+    "text": f"{text}",
+    "model_id": "eleven_monolingual_v1",
+    "voice_settings": {
+        "stability": 0.5,
+        "similarity_boost": 0.5
+    }
+    }
+
+    response = requests.post(url, json=data, headers=headers)
+    # with open('output.mp3', 'wb') as f:
+    #     for chunk in response.iter_content(chunk_size=CHUNK_SIZE):
+    #         if chunk:
+    #             f.write(chunk)
+    return "Success", response
+
+def microsoft(text, name, style="Neural"):
+    """
+    :param text:
+    :param name:
+    :param style:
+    :return:
+    """
+    headers = {
+        'Ocp-Apim-Subscription-Key': '1f1ef0ce53b84261be94fab81df7e628',
+        'Content-Type': 'application/ssml+xml',
+        'X-Microsoft-OutputFormat': 'audio-16khz-128kbitrate-mono-mp3',
+        'User-Agent': 'curl',
+    }
+
+    data = ("<speak version='1.0' xml:lang='en-US'>"
+        f"<voice xml:lang='en-US' name='{name}'>" # xml:gender='Female' 
+        f"{text}"
+        "</voice>"
+        "</speak>")
+
+    response = requests.post(
+        'https://japaneast.tts.speech.microsoft.com/cognitiveservices/v1',
+        headers=headers,
+        data=data,
+    )
+    return "Success", "sss"
+    
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--device', type=str, default='cuda')
+    parser.add_argument("--share", action="store_true", default=True, help="share gradio app")
+    parser.add_argument("--port", type=int, default=8081, help="port")
+    parser.add_argument('--model_info_path', type=str, default='/gluster/speech_data/info.json')
+    args = parser.parse_args()
+
+    app = gr.Blocks()
+    with app:
+        gr.Markdown("## Japanese TTS Demo")
+        with gr.Tabs():
+         
+            with gr.TabItem("11Labs"):
+                tts_input1 = gr.TextArea(label="Text", value=all_example)
+                tts_input2 = gr.Dropdown(eleven_name, label="name")                             
+                tts_submit = gr.Button("Generate", variant="primary")
+                tts_output1 = gr.Textbox(label="Output Message")
+                tts_output2 = gr.Audio(label="Output Audio")
+                tts_submit.click(elevenlabs, [tts_input1, tts_input2],
+                                    [tts_output1, tts_output2])
+
+            with gr.TabItem("微软"):
+                tts_input1 = gr.TextArea(label="Text", value=all_example)
+                tts_input2 = gr.Dropdown(microsoft_model_list, label="name")                             
+                tts_submit = gr.Button("Generate", variant="primary")
+                tts_output1 = gr.Textbox(label="Output Message")
+                tts_output2 = gr.Audio(label="Output Audio")
+                tts_submit.click(microsoft, [tts_input1, tts_input2],
+                                    [tts_output1, tts_output2])
+                
+            with gr.TabItem("openai"):
+                tts_input1 = gr.TextArea(label="Text", value=all_example)
+                tts_input2 = gr.Dropdown(openai_model_list, label="name")
+                tts_submit = gr.Button("Generate", variant="primary")
+                tts_output1 = gr.Textbox(label="Output Message")
+                tts_output2 = gr.Audio(label="Output Audio")
+                tts_submit.click(openai, [tts_input1, tts_input2],
+                                    [tts_output1, tts_output2])
+
+    app.queue(max_size=10)
+    app.launch(share=True)
+    # _, audio = openai(all_example,'alloy')
+    # print(audio)
+    # with open("test99.mp3", "wb") as f:
+    #     f.write(audio.content)
+