mrq b640fabab5 borrowed muon since it might better work under deepspeed and not require cruft (even though it really does not like the masked-NAR, also make the masked-NAR faux-causal since it might better help out for cfg.model.version >= 7		2025-02-23 17:23:24 -06:00
..
arch	added muon optimizer through kludge hacks because it necessitates a second optimizer in tandum that seems to only sometimes work with deepspeed	2025-02-23 11:22:13 -06:00
__init__.py		2024-12-26 21:42:17 -06:00
ar_nar.py	borrowed muon since it might better work under deepspeed and not require cruft (even though it really does not like the masked-NAR, also make the masked-NAR faux-causal since it might better help out for cfg.model.version >= 7	2025-02-23 17:23:24 -06:00
base.py	borrowed muon since it might better work under deepspeed and not require cruft (even though it really does not like the masked-NAR, also make the masked-NAR faux-causal since it might better help out for cfg.model.version >= 7	2025-02-23 17:23:24 -06:00
lora.py	naive model offloading support (handles automatically splitting parts of the model to requested device per memory constraints, either inferred or requested in the yaml, input tensors are automatically migrated to the right device, it SEEMS to work for training under the test trainer when split between GPU and CPU) (this was specifically only because that Flux imagegen model released so I can test it there)	2024-08-01 20:12:06 -05:00