Spaces:
Running on Zero
Encountering errors when attempting to generate speech
Hello; thank you for this wonderful space! I am encountering errors generating speech, no matter if I use custom audio or not. I am careful to only be typing a sentence or two at a time for input. Is there something I should be doing differently in order to get a successful generation? Any help with this is greatly appreciated by me.
Same, problem.
Okay, glad it isn't just me. Clearly this space is half baked and not ready to go.
I have this problem too
Dear owner of the space, please fix this when possible, as this is the only higgs audio v3 zerogpu space! (or maybe someone else could make their own 🤔 - and if someone does make their own, please put the link in this discussion!!)
I have this problem too
Wait a moment... there's a lemon emoji on the name of the space. It really is a lemon! At least for now. This is the third space by Multimodalart which I am experiencing extreme difficulty using. The other two are Miso TTS and Khala.
I have forked it and fixed GPU support, now it works. Also I have changed asr to parakeet v3.
https://huggingface.co/spaces/patriotyk/higgs-audio-v3-tts
Wow, thank you so much for forking it! As the Multimodalart space is a non-functional abandoned space, I am very thankful for your fork. It works! I just have to use references of .wav files instead of M4A, but it works!