Skip to content

Releases: TuringLang/Turing.jl

v0.27.1

29 Jul 19:41
fa3a6c1
Compare
Choose a tag to compare

Turing v0.27.1

Diff since v0.27.0

Closed issues:

  • Save samples as we go (#2015)

Merged pull requests:

v0.27.0

26 Jul 17:46
75961e7
Compare
Choose a tag to compare

Turing v0.27.0

Diff since v0.26.6

Closed issues:

  • Different order of parameters in MCMC chain vs. MAP result (#1322)
  • Allow silencing variational inference log messages (#1493)
  • Any plans on implementing DREAM (Differential Evolution based particle inference)? (#1640)
  • Kernel dies with Gibbs or ADVI with models involving categorical or dirichlet (#1777)
  • Feature request: Output maximized ELBO after variational inference (#1896)
  • stack overflow in ADVI (#1910)

Merged pull requests:

v0.26.6

18 Jul 22:57
e32bb71
Compare
Choose a tag to compare

Turing v0.26.6

Diff since v0.26.5

Merged pull requests:

v0.26.5

17 Jul 18:52
e1d3fef
Compare
Choose a tag to compare

Turing v0.26.5

Diff since v0.26.4

Merged pull requests:

v0.26.4

12 Jul 16:53
59e5cce
Compare
Choose a tag to compare

Turing v0.26.4

Diff since v0.26.3

Closed issues:

  • coeftable on optimize output (MLE and MAP) throws error (#2033)

Merged pull requests:

v0.26.3

10 Jul 16:13
90e1d21
Compare
Choose a tag to compare

Turing v0.26.3

Diff since v0.26.2

Merged pull requests:

v0.26.2

21 Jun 11:06
3e8d97f
Compare
Choose a tag to compare

Turing v0.26.2

Diff since v0.26.1

Merged pull requests:

v0.26.1

19 Jun 15:41
970abac
Compare
Choose a tag to compare

Turing v0.26.1

Diff since v0.26.0

Closed issues:

  • predict() from a Prior-sampled chain returns no predictions (#2012)

Merged pull requests:

v0.26.0

15 Jun 12:11
6a307cd
Compare
Choose a tag to compare

Turing v0.26.0

Diff since v0.25.3

Closed issues:

  • Implement "Variational Autoencoder" with ADVI (#1885)

Merged pull requests:

v0.25.3

13 Jun 23:10
f8450d6
Compare
Choose a tag to compare

Turing v0.25.3

Diff since v0.25.2

Merged pull requests:

  • Warn user if we're struggling to find good init params (#1999) (@torfjelde)