AIForce/train-dalle-together @ azure

暫無描述

Aleksandr Borzunov da979f7079 tweak args		3 年之前
lib	64dee420da Upgrade to using hivemind.optim.experimental	3 年之前
.gitignore	72fc0bcdb7 Initial commit (ru-max branch without private code)	3 年之前
README.md	72fc0bcdb7 Initial commit (ru-max branch without private code)	3 年之前
arguments.py	da979f7079 tweak args	3 年之前
callback.py	64dee420da Upgrade to using hivemind.optim.experimental	3 年之前
data.py	c61c61b20d Use t5-small tokenizer	3 年之前
huggingface_auth.py	72fc0bcdb7 Initial commit (ru-max branch without private code)	3 年之前
manage_scaleset.py	2816c5c09a tweak params	3 年之前
requirements.txt	64dee420da Upgrade to using hivemind.optim.experimental	3 年之前
run_aux_peer.py	64dee420da Upgrade to using hivemind.optim.experimental	3 年之前
run_trainer.py	2816c5c09a tweak params	3 年之前
run_trainer_tpu.py	64dee420da Upgrade to using hivemind.optim.experimental	3 年之前
task.py	2816c5c09a tweak params	3 年之前
utils.py	f621362466 Make logging less verbose	3 年之前

TODO write a setup instruction

Note: You might want to have several initial peers so that if one dies,

new workers still can join the collaboration via alive initial peers' addresses.
Specify initial_peers argument for that purpose