mrq 26fbb92ec6 reduced dynamic temperature threshold to > 1.0, as it seems to not quite be useful for audio LMs, sped up any sampling that touches logits by copying them to CPU first, as accessing tensors on the GPU is slow as balls)		2023-10-09 14:46:17 -05:00
..
emb
engines	fixed training stats not loading from exported weights, a bit of a readme cleanup, updated example training yaml	2023-09-23 19:59:00 -05:00
models	reduced dynamic temperature threshold to > 1.0, as it seems to not quite be useful for audio LMs, sped up any sampling that touches logits by copying them to CPU first, as accessing tensors on the GPU is slow as balls)	2023-10-09 14:46:17 -05:00
utils	implicitly load checkpoint if deepspeed checkpoint not found, updated setup script to grab the diskcached dataloader things	2023-10-06 10:02:45 -05:00
__init__.py
__main__.py	added mirostat sampling (given a partially trained model, it got far decent output than I expected, need to test on a better trained model)	2023-09-18 18:55:41 -05:00
config.py	added compat flags for torchscale because the maintainer for torchscale broke compat for existing models	2023-10-05 16:39:46 -05:00
data.py	disabled preparing of SpeechX tasks, added dynamic temperature testing (to-do: test it, credited in the function)	2023-10-09 13:01:40 -05:00
export.py
inference.py	reduced dynamic temperature threshold to > 1.0, as it seems to not quite be useful for audio LMs, sped up any sampling that touches logits by copying them to CPU first, as accessing tensors on the GPU is slow as balls)	2023-10-09 14:46:17 -05:00
plot.py	added min-x and min-y arguments to plot.py, helper script to download from my existing checkpoint	2023-10-04 19:41:37 -05:00
train.py	revert the frankensteined "train one model but hotload the other" since it kept loading the last exported weights and I'm not supporting this usecase anymore anyways	2023-09-22 13:04:17 -05:00
webui.py	reduced dynamic temperature threshold to > 1.0, as it seems to not quite be useful for audio LMs, sped up any sampling that touches logits by copying them to CPU first, as accessing tensors on the GPU is slow as balls)	2023-10-09 14:46:17 -05:00