Neural Network Pruning - Revision history

BPeat at 18:38, 12 May 2019

2019-05-12T18:38:29Z

BPeat at 18:36, 12 May 2019

2019-05-12T18:36:11Z

BPeat at 18:34, 12 May 2019

2019-05-12T18:34:09Z

BPeat at 18:23, 12 May 2019

2019-05-12T18:23:54Z

BPeat at 18:10, 12 May 2019

2019-05-12T18:10:10Z

BPeat at 18:06, 12 May 2019

2019-05-12T18:06:41Z

BPeat at 18:01, 12 May 2019

2019-05-12T18:01:01Z

BPeat at 17:59, 12 May 2019

2019-05-12T17:59:32Z

BPeat: BPeat moved page Lottery Ticket Hypothesis to Neural Network Pruning without leaving a redirect

2019-05-12T17:44:04Z

BPeat moved page Lottery Ticket Hypothesis to Neural Network Pruning without leaving a redirect

BPeat at 17:41, 12 May 2019

2019-05-12T17:41:39Z

@@ Line 9: / Line 9: @@
 * [http://openreview.net/pdf?id=rJl-b3RcF7 The Lottery Ticket Hypothesis: Finfing Spare, Trainable Neural Networks | Jonathan Frankle & Michael Carbin - MIT CSAIL]
 * [http://www.degruyter.com/downloadpdf/j/comp.2013.3.issue-3/s13537-013-0109-x/s13537-013-0109-x.pdf Pruning algorithms of neural networks - a comparative study | M. Gethsiyal Augasta, T. Kathirvalavakumar]
 * [http://github.com/tensorflow/tensorflow/tree/master/tensorflow/contrib/model_pruning Model pruning: Training tensorflow models to have masked connections]

@@ Line 19: / Line 19: @@
 <youtube>3yOZxmlBG3Y</youtube>
 <youtube>zTzBSYdXKxk</youtube>
-<youtube>3yOZxmlBG3Y</youtube>
+<youtube>vouEMwDNopQ</youtube>

← Older revision		Revision as of 18:34, 12 May 2019
Line 17:		Line 17:

	<youtube>s7DqRZVvRiQ</youtube>		<youtube>s7DqRZVvRiQ</youtube>
		+	<youtube>3yOZxmlBG3Y</youtube>
		+	<youtube>zTzBSYdXKxk</youtube>
		+	<youtube>3yOZxmlBG3Y</youtube>

@@ Line 10: / Line 10: @@
 * [http://openreview.net/pdf?id=rJl-b3RcF7 The Lottery Ticket Hypothesis: Finfing Spare, Trainable Neural Networks | Jonathan Frankle & Michael Carbin - MIT CSAIL]
 * [http://www.degruyter.com/downloadpdf/j/comp.2013.3.issue-3/s13537-013-0109-x/s13537-013-0109-x.pdf Pruning algorithms of neural networks - a comparative study | M. Gethsiyal Augasta, T. Kathirvalavakumar]
 Lottery Ticket Hypothesis: The discovery hinges on the reality that the random connection strengths assigned during initialization aren’t, in fact, random in their consequences: they predispose different parts of the network to fail or succeed before training even happens. Put another way, the initial configuration influences which final configuration the network will arrive at. By focusing on this idea, the researchers found that if you prune an oversize network after training, you can actually reuse the resultant smaller network to train on new data and preserve high performance—as long as you reset each connection within this downsized network back to its initial strength.

@@ Line 5: / Line 5: @@
 |description=Helpful resources for your journey with artificial intelligence; videos, articles, techniques, courses, profiles, and tools
 }}
-[http://www.youtube.com/results?search_query=Neural+Network+Pruning+Lottery+Ticket+Hypothesis+artificial+intelligence Youtube search...]
+[http://www.youtube.com/results?search_query=Neural+Network+Pruning+artificial+intelligence Youtube search...]
-[http://www.google.com/search?q=Neural+Network+Pruning+Lottery+Ticket+Hypothesis+artificial+intelligence ...Google search]
+[http://www.google.com/search?q=Neural+Network+Pruning+artificial+intelligence ...Google search]
 * [http://openreview.net/pdf?id=rJl-b3RcF7 The Lottery Ticket Hypothesis: Finfing Spare, Trainable Neural Networks | Jonathan Frankle & Michael Carbin - MIT CSAIL]
-* [http://www.degruyter.com/downloadpdf/j/comp.2013.3.issue-3/s13537-013-0109-x/s13537-013-0109-x.pdf Pruning algorithms of neural networks - a comparative study | M. Gethsiyal Augasta
+* [http://www.degruyter.com/downloadpdf/j/comp.2013.3.issue-3/s13537-013-0109-x/s13537-013-0109-x.pdf Pruning algorithms of neural networks - a comparative study | M. Gethsiyal Augasta, T. Kathirvalavakumar]
 Lottery Ticket Hypothesis: The discovery hinges on the reality that the random connection strengths assigned during initialization aren’t, in fact, random in their consequences: they predispose different parts of the network to fail or succeed before training even happens. Put another way, the initial configuration influences which final configuration the network will arrive at. By focusing on this idea, the researchers found that if you prune an oversize network after training, you can actually reuse the resultant smaller network to train on new data and preserve high performance—as long as you reset each connection within this downsized network back to its initial strength.

← Older revision		Revision as of 18:06, 12 May 2019
Line 9:		Line 9:

	* [http://openreview.net/pdf?id=rJl-b3RcF7 The Lottery Ticket Hypothesis: Finfing Spare, Trainable Neural Networks \| Jonathan Frankle & Michael Carbin - MIT CSAIL]		* [http://openreview.net/pdf?id=rJl-b3RcF7 The Lottery Ticket Hypothesis: Finfing Spare, Trainable Neural Networks \| Jonathan Frankle & Michael Carbin - MIT CSAIL]
		+	* [http://www.degruyter.com/downloadpdf/j/comp.2013.3.issue-3/s13537-013-0109-x/s13537-013-0109-x.pdf Pruning algorithms of neural networks - a comparative study \| M. Gethsiyal Augasta
		+	, T. Kathirvalavakumar]

	Lottery Ticket Hypothesis: The discovery hinges on the reality that the random connection strengths assigned during initialization aren’t, in fact, random in their consequences: they predispose different parts of the network to fail or succeed before training even happens. Put another way, the initial configuration influences which final configuration the network will arrive at. By focusing on this idea, the researchers found that if you prune an oversize network after training, you can actually reuse the resultant smaller network to train on new data and preserve high performance—as long as you reset each connection within this downsized network back to its initial strength.		Lottery Ticket Hypothesis: The discovery hinges on the reality that the random connection strengths assigned during initialization aren’t, in fact, random in their consequences: they predispose different parts of the network to fail or succeed before training even happens. Put another way, the initial configuration influences which final configuration the network will arrive at. By focusing on this idea, the researchers found that if you prune an oversize network after training, you can actually reuse the resultant smaller network to train on new data and preserve high performance—as long as you reset each connection within this downsized network back to its initial strength.

← Older revision	Revision as of 17:44, 12 May 2019
(No difference)

← Older revision		Revision as of 17:41, 12 May 2019
Line 14:		Line 14:


−	<youtube>~~JXQT_vxqwIs~~</youtube>	+	<youtube>s7DqRZVvRiQ</youtube>