mrq 4d75ee066c actually do the Linear replacement with TE's Linear		2024-04-09 14:41:13 -05:00
..
emb
engines	actually do the Linear replacement with TE's Linear	2024-04-09 14:41:13 -05:00
ext	added FP8 support through NVIDIA/TransformerEngine, added RetNet_HF through syncdoth/RetNet (as an alternative to branch away from torchscale)	2024-04-08 20:14:51 -05:00
models	actually do the Linear replacement with TE's Linear	2024-04-09 14:41:13 -05:00
utils	actually do the Linear replacement with TE's Linear	2024-04-09 14:41:13 -05:00
__init__.py
__main__.py
config.py	added FP8 support through NVIDIA/TransformerEngine, added RetNet_HF through syncdoth/RetNet (as an alternative to branch away from torchscale)	2024-04-08 20:14:51 -05:00
data.py
export.py
inference.py
plot.py
samplers.py
train.py	logger broke for some reason, added flag to just tqdm.write instead, make cfg.bitsandbytes.bitnet==True yamls denoted since I'm sure they're not interoperable	2024-03-01 10:32:35 -06:00
webui.py