Joe Rocca's picture

Joe Rocca

rocca

·

rocca27

AI & ML interests

None yet

Recent Activity

liked a model about 6 hours ago

Alpha-VLLM/Lumina-Image-2.0

reacted to hexgrad's post with 👍 3 days ago

I wrote an article about G2P: https://hf.co/blog/hexgrad/g2p G2P is an underrated piece of small TTS models, like offensive linemen who do a bunch of work and get no credit. Instead of relying on explicit G2P, larger speech models implicitly learn this task by eating many thousands of hours of audio data. They often use a 500M+ parameter LLM at the front to predict latent audio tokens over a learned codebook, then decode these tokens into audio. Kokoro instead relies on G2P preprocessing, is 82M parameters, and thus needs less audio to learn. Because of this, we can cherrypick high fidelity audio for training data, and deliver solid speech for those voices. In turn, this excellent audio quality & lack of background noise helps explain why Kokoro is very competitive in single-voice TTS Arenas.

liked a Space 5 days ago

lllyasviel/iclight-v2

View all activity

Organizations

rocca's activity

New activity in WildVision/vision-arena 7 months ago

"NETWORK ERROR DUE TO HIGH TRAFFIC. PLEASE REGENERATE OR REFRESH THIS PAGE." - votes should be ignored in this case

#13 opened 7 months ago by

New activity in ibm-ai-platform/llama-13b-accelerator 7 months ago

70B model?

#1 opened 8 months ago by

New activity in MaziyarPanahi/miqu-1-70b-sf-GPTQ 8 months ago

vllm can not inter this model (other 70b gptq model are ok)

#1 opened about 1 year ago by

New activity in text-generation-inference/commandrplus-medusa 10 months ago

Is `--speculate` needed for Medusa in TGI?

#1 opened 10 months ago by

New activity in TheBloke/goliath-120b-GPTQ about 1 year ago

Double-hyphen in "gptq-3bit--1g-actorder_True" branch name is invalid according to TGI

#1 opened about 1 year ago by

New activity in TheBloke/LLaMA2-13B-Tiefighter-GPTQ about 1 year ago

TGI "Fast Tokenizer" support?

#1 opened about 1 year ago by

New activity in tiiuae/falcon-40b-instruct over 1 year ago

Fix "Finetuned from model" link

#26 opened over 1 year ago by

New activity in rocca/lyra-v2-soundstream about 2 years ago

Conversion to ONNX

#1 opened about 2 years ago by

New activity in rocca/rwkv-4-pile-web over 2 years ago

Delete the outdated rwkv-4-pile-430m-webgl.onnx

#4 opened over 2 years ago by

Upload rwkv-4-pile-430m-webgl.onnx

#3 opened over 2 years ago by

Upload rwkv-4-pile-430m-webgl.onnx

#2 opened over 2 years ago by

Upload rwkv-4-pile-169m-webgl.onnx

#1 opened over 2 years ago by

New activity in rocca/openai-clip-js over 2 years ago

This is really exciting

#1 opened over 2 years ago by