Commit fed3f38

committed

Update README.md

1 parent 53beaf3 commit fed3f38Copy full SHA for fed3f38

File tree

+11

-0

lines changed

+11

-0

lines changed

Lines changed: 11 additions & 0 deletions

Original file line number	Diff line number	Diff line change
`@@ -31,3 +31,14 @@ import torchbnn`
`31`	`31`	`* modules : BayesLinear, BayesConv2d, BayesBatchNorm2d`
`32`	`32`	`* utils : convert_model(nonbayes_to_bayes, bayes_to_nonbayes)`
`33`	`33`	`* functional : bayesian_kl_loss`
	`34`	`+<<<<<<< HEAD`
	`35`	`+`
	`36`	`+### Version 0.2`
	`37`	`+* prior_sigma is used when initialize modules and functions instead of prior_log_sigma`
	`38`	`+ * Modules(BayesLinear, BayesConv2d, BayesBatchNorm2d) are re-defined with prior_sigma instead of prior_log_sigma.`
	`39`	`+ * convert_model(nonbayes_to_bayes, bayes_to_nonbayes) is also changed with prior_sigma instead of prior_log_sigma.`
	`40`	`+* Modules(BayesLinear, BayesConv2d, BayesBatchNorm2d) : Base initialization method is changed to the method of Adv-BNN from the original torch method.`
	`41`	`+* functional : bayesian_kl_loss is changed similar to ones in torch.functional`
	`42`	`+* loss : BKLLoss is added based on bayesian_kl_loss similar to ones in torch.loss`
	`43`	`+=======`
	`44`	`+>>>>>>> 8be43015d291edb1fc413d4ec25e21d238374f70`

Comments

(0)