mrq 458b95d196 added option to split between text loss and audio loss (to-do: document this better), because it may or may not be a problem with LLaMA-backed models because my loss hovers around 3.9 / 56% accuracy despite sounding decent at the moment		2024-05-19 11:23:56 -05:00
..
config.yaml	added option to split between text loss and audio loss (to-do: document this better), because it may or may not be a problem with LLaMA-backed models because my loss hovers around 3.9 / 56% accuracy despite sounding decent at the moment	2024-05-19 11:23:56 -05:00
qnt.dac	god I need to replace omegaconf	2024-05-12 14:01:52 -05:00
qnt.pt
tokenizer.json	final tweaks, hopefully, again	2024-05-15 23:04:19 -05:00