Open K3NOXOFFICIAL opened 1 year ago
I think you can hardcode the API key in config.json
line 11
I think you can hardcode the API key in
config.json
line 11
I Did that already, I can see it loading it in the terminal (replaced the api keys with ------) but it isn't showing in the settings and I have to manually turn every option off and on again as well as paste the keys in again for it to work again. otherwise it doesn't work or I have no sound output.
loading config... {'use_deepl': False, 'deepl_api_key': '-------------x', 'use_cloud_voice_vox': False, 'voice_vox_api_key': '', 'openai_api_key': '------------------', 'youtube_video_id': '', 'twitch_access_token': '', 'twitch_channel_name': '', 'use_elevenlab': True, 'elevenlab_api_key': '--------------'}
Maybe you should post your console log here for Xiaohei check it.
Maybe you should post your console log here for Xiaohei check it.
`Loading... Starting voicevox server... Initializing tts model... Initializing translator... Warning: cpu_num_threads is set to 0. ( The library leaves the decision to the synthesis runtime ) 2023-06-08T13:48:18.556655Z INFO voicevox_core::publish: 検出されたGPU (DirectMLには1番目のGPUが使われます): 2023-06-08T13:48:18.560487Z INFO voicevox_core::publish: - "NVIDIA GeForce RTX 3060 Ti" (7.85 GiB) 2023-06-08T13:48:18.560537Z INFO voicevox_core::publish: - "Microsoft Basic Render Driver" (0 B) INFO: Started server process [21216] INFO: Waiting for application startup. reading C:\Users\Kenos\AppData\Local\voicevox-engine\voicevox-engine\tmppvvgs6xd ... 62 emitting double-array: 100% |###########################################|
done! INFO: Application startup complete. INFO: Uvicorn running on http://127.0.0.1:50021 (Press CTRL+C to quit) C:\AI\VoiceToJapanese.v1.2.5\UI-v1.2.5-py\files.venv\lib\site-packages\transformers\models\marian\tokenization_marian.py:194: UserWarning: Recommended: pip install sacremoses. warnings.warn("Recommended: pip install sacremoses.") loading Audio devices: loading config... {'use_deepl': False, 'deepl_api_key': '-------------x', 'use_cloud_voice_vox': False, 'voice_vox_api_key': '', 'openai_api_key': '------------------', 'youtube_video_id': '', 'twitch_access_token': '', 'twitch_channel_name': '', 'use_elevenlab': True, 'elevenlab_api_key': '--------------'} loading settings... {'output_device': 'CABLE Input (VB-Audio Virtual C', 'elevenlab_voice_name': 'Domi', 'output_audio_driver': 'MME'} INFO: 127.0.0.1:50764 - "GET /speakers HTTP/1.1" 200 OK [{'name': 'ノーマル', 'id': 2}, {'name': 'あまあま', 'id': 0}, {'name': 'ツンツン', 'id': 6}, {'name': 'セクシー', 'id': 4}, {'name': 'ささやき', 'id': 36}, {'name': 'ヒソヒソ', 'id': 37}] [{'name': 'ノーマル', 'id': 2}, {'name': 'あまあま', 'id': 0}, {'name': 'ツンツン', 'id': 6}, {'name': 'セクシー', 'id': 4}, {'name': 'ささやき', 'id': 36}, {'name': 'ヒソヒソ', 'id': 37}] setting[mic_mode] is not found. setting[push_to_talk_key] is not found. setting[input_audio_driver] is not found. Looking for in ['MME', 'Windows DirectSound', 'ASIO', 'Windows WASAPI', 'Windows WDM-KS']. setting[input_device] is not found.`
I think the issue is with it loading 'voice_vox_api_key' instead of 'elevenlab_api_key' in the Ui.py. I replaced it in my forked repo and that seemed to fix the issue.
You just need to replace it with 'elevenlab_api_key' in Ui.py for elevenlabs in the SettingsFrame class.
@anaveragecha I had a look on your repo a few days ago but couldn't get it to work. Can you work out a Release? Other than that I will try it out so thanks for the tips.
@K3NOXOFFICIAL Sure, I just made a pre-release, you can find it under releases on my forked repo. The installation is the same as 0Xiaohei0, along with the .zip file, the only difference is that I replaced 4 files in it. You can also try to replace the 4 files to avoid doing a fresh install (more info in my pre-release).
@anaveragecha Thanks a lot I will have a look at it and try it out.
Edit: Works Perfectly now after the made changes, also do you have any idea on how to use English Speech-to-text without the need of Elevenlabs because the Characters are gone really quickly and the subscriptions are really expensive.
@K3NOXOFFICIAL I was toying with the idea of using tortoise-tts for english tts a few days ago and made an alpha pre-release just now. There's more info on my wiki, the section How to use with mrq's gradio api. This is intended for locally hosting each one so it will take up loads of memory.
@anaveragecha No Worries I got loads of ram and a decent AI server if it fails on my PC so i'll try it out, really appreciate your commitment for this software. Also thanks a lot for helping the community.
P.S. till now your repo works perfectly I see no bugs tho you should create the issues tab for when there are issues or bugs btw it runs far better than the this repo no idea why
Bug Description: Upon restarting the application, the Evenlabs API key is being deleted, resulting in the inability to access Evenlabs services and functionality. This issue prevents the application from successfully communicating with the Evenlabs API, leading to not working TTS in the application.
Steps to Reproduce:
Start the application.
Expected Result: After restarting the application, the Evenlabs API key should be retained in the application settings, allowing seamless access to Evenlabs services and functionality.
Actual Result: Upon restarting the application, the Evenlabs API key is deleted from the application settings, rendering it impossible to access Evenlabs services and functionality.
Additional Information:
The issue occurs consistently on every application restart. Other API keys or settings are not affected by this problem. The application's log files do not indicate any errors or relevant information related to the deletion of the Evenlabs API key. The application is running on Windows 11.