Skip to content
Snippets Groups Projects
Commit ff482913 authored by Erick Lavoie's avatar Erick Lavoie
Browse files

Completed first draft of evaluation section

parent 437d8e52
No related branches found
No related tags found
No related merge requests found
Showing
with 221 additions and 146 deletions
figures/d-cliques-mnist-linear-comparison-to-non-clustered-topologies.png

88 KiB | W: | H:

figures/d-cliques-mnist-linear-comparison-to-non-clustered-topologies.png

90.1 KiB | W: | H:

figures/d-cliques-mnist-linear-comparison-to-non-clustered-topologies.png
figures/d-cliques-mnist-linear-comparison-to-non-clustered-topologies.png
figures/d-cliques-mnist-linear-comparison-to-non-clustered-topologies.png
figures/d-cliques-mnist-linear-comparison-to-non-clustered-topologies.png
  • 2-up
  • Swipe
  • Onion skin
...@@ -345,7 +345,8 @@ of the local models across nodes. ...@@ -345,7 +345,8 @@ of the local models across nodes.
inter-clique connections (see main text for details).} inter-clique connections (see main text for details).}
\end{figure} \end{figure}
\paragraph{Clique Averaging.} We address this problem by adding \emph{Clique \paragraph{\label{section:clique-averaging} Clique Averaging.}
We address this problem by adding \emph{Clique
Averaging} to D-SGD Averaging} to D-SGD
(Algorithm~\ref{Algorithm:Clique-Unbiased-D-PSGD}), which essentially (Algorithm~\ref{Algorithm:Clique-Unbiased-D-PSGD}), which essentially
decouples gradient averaging from model averaging. The idea is to use only the decouples gradient averaging from model averaging. The idea is to use only the
......
This diff is collapsed.
mlsys2022style/figures/convergence-cifar10-random-vs-d-cliques-2-shards.png

50.3 KiB | W: | H:

mlsys2022style/figures/convergence-cifar10-random-vs-d-cliques-2-shards.png

53 KiB | W: | H:

mlsys2022style/figures/convergence-cifar10-random-vs-d-cliques-2-shards.png
mlsys2022style/figures/convergence-cifar10-random-vs-d-cliques-2-shards.png
mlsys2022style/figures/convergence-cifar10-random-vs-d-cliques-2-shards.png
mlsys2022style/figures/convergence-cifar10-random-vs-d-cliques-2-shards.png
  • 2-up
  • Swipe
  • Onion skin
mlsys2022style/figures/convergence-mnist-random-vs-d-cliques-2-shards.png

55.8 KiB | W: | H:

mlsys2022style/figures/convergence-mnist-random-vs-d-cliques-2-shards.png

60.5 KiB | W: | H:

mlsys2022style/figures/convergence-mnist-random-vs-d-cliques-2-shards.png
mlsys2022style/figures/convergence-mnist-random-vs-d-cliques-2-shards.png
mlsys2022style/figures/convergence-mnist-random-vs-d-cliques-2-shards.png
mlsys2022style/figures/convergence-mnist-random-vs-d-cliques-2-shards.png
  • 2-up
  • Swipe
  • Onion skin
mlsys2022style/figures/convergence-speed-cifar10-dc-fc-vs-fc-2-shards-per-node.png

43.9 KiB | W: | H:

mlsys2022style/figures/convergence-speed-cifar10-dc-fc-vs-fc-2-shards-per-node.png

44.4 KiB | W: | H:

mlsys2022style/figures/convergence-speed-cifar10-dc-fc-vs-fc-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-cifar10-dc-fc-vs-fc-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-cifar10-dc-fc-vs-fc-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-cifar10-dc-fc-vs-fc-2-shards-per-node.png
  • 2-up
  • Swipe
  • Onion skin
mlsys2022style/figures/convergence-speed-cifar10-dc-random-vs-dc-gs-2-shards-per-node.png

44.6 KiB | W: | H:

mlsys2022style/figures/convergence-speed-cifar10-dc-random-vs-dc-gs-2-shards-per-node.png

46.3 KiB | W: | H:

mlsys2022style/figures/convergence-speed-cifar10-dc-random-vs-dc-gs-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-cifar10-dc-random-vs-dc-gs-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-cifar10-dc-random-vs-dc-gs-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-cifar10-dc-random-vs-dc-gs-2-shards-per-node.png
  • 2-up
  • Swipe
  • Onion skin
mlsys2022style/figures/convergence-speed-cifar10-w-c-avg-no-mom-vs-mom-2-shards-per-node.png

46.6 KiB | W: | H:

mlsys2022style/figures/convergence-speed-cifar10-w-c-avg-no-mom-vs-mom-2-shards-per-node.png

51.3 KiB | W: | H:

mlsys2022style/figures/convergence-speed-cifar10-w-c-avg-no-mom-vs-mom-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-cifar10-w-c-avg-no-mom-vs-mom-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-cifar10-w-c-avg-no-mom-vs-mom-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-cifar10-w-c-avg-no-mom-vs-mom-2-shards-per-node.png
  • 2-up
  • Swipe
  • Onion skin
mlsys2022style/figures/convergence-speed-cifar10-wo-c-avg-no-mom-vs-mom-2-shards-per-node.png

45.3 KiB | W: | H:

mlsys2022style/figures/convergence-speed-cifar10-wo-c-avg-no-mom-vs-mom-2-shards-per-node.png

50.5 KiB | W: | H:

mlsys2022style/figures/convergence-speed-cifar10-wo-c-avg-no-mom-vs-mom-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-cifar10-wo-c-avg-no-mom-vs-mom-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-cifar10-wo-c-avg-no-mom-vs-mom-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-cifar10-wo-c-avg-no-mom-vs-mom-2-shards-per-node.png
  • 2-up
  • Swipe
  • Onion skin
mlsys2022style/figures/convergence-speed-mnist-dc-fc-vs-fc-2-shards-per-node.png

44.6 KiB | W: | H:

mlsys2022style/figures/convergence-speed-mnist-dc-fc-vs-fc-2-shards-per-node.png

45.7 KiB | W: | H:

mlsys2022style/figures/convergence-speed-mnist-dc-fc-vs-fc-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-mnist-dc-fc-vs-fc-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-mnist-dc-fc-vs-fc-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-mnist-dc-fc-vs-fc-2-shards-per-node.png
  • 2-up
  • Swipe
  • Onion skin
mlsys2022style/figures/convergence-speed-mnist-dc-no-c-avg-vs-c-avg-2-shards-per-node.png

62 KiB | W: | H:

mlsys2022style/figures/convergence-speed-mnist-dc-no-c-avg-vs-c-avg-2-shards-per-node.png

69.7 KiB | W: | H:

mlsys2022style/figures/convergence-speed-mnist-dc-no-c-avg-vs-c-avg-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-mnist-dc-no-c-avg-vs-c-avg-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-mnist-dc-no-c-avg-vs-c-avg-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-mnist-dc-no-c-avg-vs-c-avg-2-shards-per-node.png
  • 2-up
  • Swipe
  • Onion skin
mlsys2022style/figures/convergence-speed-mnist-dc-random-vs-dc-gs-2-shards-per-node.png

56.5 KiB | W: | H:

mlsys2022style/figures/convergence-speed-mnist-dc-random-vs-dc-gs-2-shards-per-node.png

60.4 KiB | W: | H:

mlsys2022style/figures/convergence-speed-mnist-dc-random-vs-dc-gs-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-mnist-dc-random-vs-dc-gs-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-mnist-dc-random-vs-dc-gs-2-shards-per-node.png
mlsys2022style/figures/convergence-speed-mnist-dc-random-vs-dc-gs-2-shards-per-node.png
  • 2-up
  • Swipe
  • Onion skin
mlsys2022style/figures/d-cliques-cifar10-w-clique-avg-impact-of-edge-removal.png

60.9 KiB | W: | H:

mlsys2022style/figures/d-cliques-cifar10-w-clique-avg-impact-of-edge-removal.png

65 KiB | W: | H:

mlsys2022style/figures/d-cliques-cifar10-w-clique-avg-impact-of-edge-removal.png
mlsys2022style/figures/d-cliques-cifar10-w-clique-avg-impact-of-edge-removal.png
mlsys2022style/figures/d-cliques-cifar10-w-clique-avg-impact-of-edge-removal.png
mlsys2022style/figures/d-cliques-cifar10-w-clique-avg-impact-of-edge-removal.png
  • 2-up
  • Swipe
  • Onion skin
mlsys2022style/figures/d-cliques-cifar10-wo-clique-avg-impact-of-edge-removal.png

52.4 KiB | W: | H:

mlsys2022style/figures/d-cliques-cifar10-wo-clique-avg-impact-of-edge-removal.png

56.4 KiB | W: | H:

mlsys2022style/figures/d-cliques-cifar10-wo-clique-avg-impact-of-edge-removal.png
mlsys2022style/figures/d-cliques-cifar10-wo-clique-avg-impact-of-edge-removal.png
mlsys2022style/figures/d-cliques-cifar10-wo-clique-avg-impact-of-edge-removal.png
mlsys2022style/figures/d-cliques-cifar10-wo-clique-avg-impact-of-edge-removal.png
  • 2-up
  • Swipe
  • Onion skin
mlsys2022style/figures/d-cliques-mnist-linear-comparison-to-non-clustered-topologies.png

90.1 KiB

mlsys2022style/figures/d-cliques-mnist-w-clique-avg-impact-of-edge-removal.png

72.4 KiB | W: | H:

mlsys2022style/figures/d-cliques-mnist-w-clique-avg-impact-of-edge-removal.png

78.2 KiB | W: | H:

mlsys2022style/figures/d-cliques-mnist-w-clique-avg-impact-of-edge-removal.png
mlsys2022style/figures/d-cliques-mnist-w-clique-avg-impact-of-edge-removal.png
mlsys2022style/figures/d-cliques-mnist-w-clique-avg-impact-of-edge-removal.png
mlsys2022style/figures/d-cliques-mnist-w-clique-avg-impact-of-edge-removal.png
  • 2-up
  • Swipe
  • Onion skin
mlsys2022style/figures/d-cliques-mnist-wo-clique-avg-impact-of-edge-removal.png

85 KiB | W: | H:

mlsys2022style/figures/d-cliques-mnist-wo-clique-avg-impact-of-edge-removal.png

91.8 KiB | W: | H:

mlsys2022style/figures/d-cliques-mnist-wo-clique-avg-impact-of-edge-removal.png
mlsys2022style/figures/d-cliques-mnist-wo-clique-avg-impact-of-edge-removal.png
mlsys2022style/figures/d-cliques-mnist-wo-clique-avg-impact-of-edge-removal.png
mlsys2022style/figures/d-cliques-mnist-wo-clique-avg-impact-of-edge-removal.png
  • 2-up
  • Swipe
  • Onion skin
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment