As the title says.
I'm playing around on my PC and found that it seems the API for loading a model only allows to use the Qwen models for that.
I don't know if it would be different if I pre-load say Kokoro on Voicebox directly but if I try to go over things via API then when I load a qwen model it works just fine.
But when I try to load say Kokoro or a Chatterbox one it fails:
TTS error: Voicebox API error 422: {"detail":[{"type":"string_pattern_mismatch","loc":["body","model_size"],"msg":"String should match pattern '^(1\.7B|0\.6B|1B|3B)$'","input":"chatterbox-turbo","ctx":{"pattern":"^(1\.7B|0\.6B|1B|3
Seems a bit self explanatory when it comes to the error.
Maybe I am using the wrong API call here for this model or something?
Any kind of help would be awesome.
As the title says.
I'm playing around on my PC and found that it seems the API for loading a model only allows to use the Qwen models for that.
I don't know if it would be different if I pre-load say Kokoro on Voicebox directly but if I try to go over things via API then when I load a qwen model it works just fine.
But when I try to load say Kokoro or a Chatterbox one it fails:
TTS error: Voicebox API error 422: {"detail":[{"type":"string_pattern_mismatch","loc":["body","model_size"],"msg":"String should match pattern '^(1\.7B|0\.6B|1B|3B)$'","input":"chatterbox-turbo","ctx":{"pattern":"^(1\.7B|0\.6B|1B|3
Seems a bit self explanatory when it comes to the error.
Maybe I am using the wrong API call here for this model or something?
Any kind of help would be awesome.