Narrator saying person 1 greater than

[nkenna]

We are using this library in our project and have observed a few issues we'd like to highlight.

In some of the generated audio files, it repeatedly mentions "person 1 greater than," which it shouldn't. To provide more clarity, I will attach a sample audio file link for your reference.

https://drive.google.com/file/d/1N9CLeW_yoEH1EOfAdlTMVGhfMqL5lBpL/view?usp=sharing

[souzatharsis]

Hi thanks for raising this issue. What's the tss model and llm used?

…

On Sat, Dec 28, 2024, 9:29 AM Ugwumsinachi Nnadi ***@***.***> wrote: We are using this library in our project and have observed a few issues we'd like to highlight. In some of the generated audio files, it repeatedly mentions "person 1 greater than," which it shouldn't. To provide more clarity, I will attach a sample audio file link for your reference. https://drive.google.com/file/d/1N9CLeW_yoEH1EOfAdlTMVGhfMqL5lBpL/view?usp=sharing — Reply to this email directly, view it on GitHub <#218>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADTMY3O6FS762SCX4JRYZJD2H2KRBAVCNFSM6AAAAABUJYNWKOVHI2DSMVQWIX3LMV43ASLTON2WKOZSG43DCNZQGE3DKOA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[nkenna]

Hi thanks for raising this issue. What's the tss model and llm used?
…
On Sat, Dec 28, 2024, 9:29 AM Ugwumsinachi Nnadi @.> wrote: We are using this library in our project and have observed a few issues we'd like to highlight. In some of the generated audio files, it repeatedly mentions "person 1 greater than," which it shouldn't. To provide more clarity, I will attach a sample audio file link for your reference. https://drive.google.com/file/d/1N9CLeW_yoEH1EOfAdlTMVGhfMqL5lBpL/view?usp=sharing — Reply to this email directly, view it on GitHub <#218>, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADTMY3O6FS762SCX4JRYZJD2H2KRBAVCNFSM6AAAAABUJYNWKOVHI2DSMVQWIX3LMV43ASLTON2WKOZSG43DCNZQGE3DKOA . You are receiving this because you are subscribed to this thread.Message ID: @.>

tts model = gemini
llm = gemini-1.5-pro-latest

audio_file = generate_podcast( text=content, llm_model_name="gemini-1.5-pro-latest", api_key_label="GEMINI_API_KEY", tts_model="gemini", conversation_config=conversation_config, longform=False )

[nkenna]

Good day, any update on this. It seems to be occuring regularly now.

[souzatharsis]

Not yet; it's New Year's eve :) Also, feel free to push a fix if you have a solution in mind. Happy New Year! Best Regards,

…

-- Thársis <http://linkedin.com/in/tharsissouza>

On Tue, Dec 31, 2024 at 1:38 PM Ugwumsinachi Nnadi ***@***.***> wrote: Good day, any update on this. It seems to be occuring regularly now. — Reply to this email directly, view it on GitHub <#218 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADTMY3JL5HFVMCXEJSMWYKT2ILCBPAVCNFSM6AAAAABUJYNWKOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKNRWGU4DAOJZHE> . You are receiving this because you commented.Message ID: ***@***.***>

[souzatharsis]

Hi, do you happen to have the generated transcript?

I am trying to figure out whether the issue is in the transcript generation or text-to-speech processing.

Thanks!

[nkenna]

Sorry. The transcript files are deleted after the audio generation but I will look out for another and keep the file.

[nkenna]

transcript_5a84f0c73e3e4f57a4b0938e27feb633.txt

[nkenna]

transcript_5a84f0c73e3e4f57a4b0938e27feb633.txt

Check the last line