You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

Super Daniel 0dbd61c29b [fx] fix test and algorithm bugs in activation checkpointing. (#1451) ... * [fx] modify the calculation of node_size in MetaInfoProp for activation checkpointing usages * [fx] modify the calculation of node_size in MetaInfoProp for activation checkpointing usages * [fx] modify the calculation of node_size in MetaInfoProp for activation checkpointing usages * [fx] merge development into main (#1) * [fx] activation checkpointing using Chen strategies. * [fx] add test for ckpt_solver_chen * [fx] add vanilla activation checkpoint search with test on resnet and densenet * [fx] add a namespace code for solver_chen. * [fx] fix the false interpretation of algorithm 3 in https://arxiv.org/abs/1604.06174. * [fx] fix lowercase naming conventions. * [fx] simplify test for ckpt. * [fx] fix test and algorithm bugs in activation checkpointing. * mend [fx] fix test and algorithm bugs in activation checkpointing. * mend [fx] fix test and algorithm bugs in activation checkpointing. * mend [fx] fix test and algorithm bugs in activation checkpointing. * mend [fx] fix test and algorithm bugs in activation checkpointing. * [fx] polish ckpt_test. * [fx] polish ckpt_test. * [fx] polish ckpt_test.		2 years ago
..
amp	[doc] update rst and docstring (#1351)	2 years ago
builder	…
cli	…
communication	[communication] add p2p_v2.py to support communication with List[Any] (#1407)	2 years ago
context	[doc] update rst and docstring (#1351)	2 years ago
device	[device] add DeviceMesh class to support logical device layout (#1394)	2 years ago
engine	[engin/schedule] use p2p_v2 to recontruct pipeline_schedule (#1408)	2 years ago
fx	[fx] fix test and algorithm bugs in activation checkpointing. (#1451)	2 years ago
gemini	[zero] add chunk_managerV2 for all-gather chunk (#1441)	2 years ago
kernel	[hotfix] fix CPUAdam kernel nullptr (#1410)	2 years ago
logging	…
nn	fix nvme docstring (#1450)	2 years ago
pipeline	…
registry	…
tensor	[tensor] shape consistency generate transform path and communication cost (#1435)	2 years ago
testing	…
trainer	…
utils	[test] fixed the activation codegen test (#1447)	2 years ago
zero	[utils] Impl clip_grad_norm for ColoTensor and ZeroOptimizer (#1442)	2 years ago
__init__.py	…
constants.py	…
core.py	…
global_variables.py	…
initialize.py	…