mirror of
https://github.com/hpcaitech/ColossalAI.git
synced 2025-09-04 18:40:28 +00:00
[fx] provide a stable but not accurate enough version of profiler. (#1547)
* [fx] compute memory stat and flop count for MetaInfoProp. * [fx] modify node attribute. * [fx] modify ckpt_chen. * [fx] fix compatibility. * [fx] fix import error. * [fx] skip test for MetaInfoProp. * [fx] skip test for MetaInfoProp. * [fx] skip test for MetaInfoProp. * [fx] skip test for MetaInfoProp. * [fx] skip if torch 1.11.0. * [fx] recover MetaInfoProp support for PyTorch 1.11. * [fx] provide a stable but not accurate enough version of profiler. * [fx] provide a stable but not accurate enough version of profiler. * [fx] fix compatibility in tests. * [fx] fix compatibility in tests. * [fx] fix compatibility in tests. * [fx] fix compatibility in tests. * [fx] fix compatibility in tests. * [fx] fix compatibility in tests. * [fx] fix compatibility in tests. * [fx] fix compatibility in tests. * [fx] fix compatibility in tests. * [fx] fix compatibility in tests. * [fx] fix import error.
This commit is contained in:
@@ -73,10 +73,10 @@ def chen_greedy(gm: GraphModule) -> GraphModule:
|
||||
y = 0
|
||||
prev_idx = 2
|
||||
for (idx, n) in enumerate(gm.graph.nodes):
|
||||
temp += getattr(n, '__activation__')
|
||||
temp += getattr(n, 'fwd_out')
|
||||
y = max(y, temp)
|
||||
if temp > b and n in ckpt_nodes:
|
||||
x += getattr(n, '__activation__')
|
||||
x += getattr(n, 'fwd_out')
|
||||
temp = 0
|
||||
ckpt_intv.append((prev_idx, idx + 1))
|
||||
prev_idx = idx + 1
|
||||
|
Reference in New Issue
Block a user