15 Commits

Author	SHA1	Message	Date
iko	afb9f9bca5	Updated to lts-20.18	2023-04-22 18:02:24 +03:00
iko	b79918bf0f	Bumped to lts-18.28	2022-03-22 17:24:02 +03:00
Erik de Castro Lopo	5206c95c42	Remove use of RecordWildCards	2020-04-12 06:24:42 +10:00
Erik de Castro Lopo	14ec0de6bf	examples: Use CPP to support multiple GHC versions	2020-01-27 08:00:41 +11:00
Erik de Castro Lopo	fed6eb2c70	Drop support for GHC < 8.0	2020-01-19 14:59:00 +11:00
Erik de Castro Lopo	2b5dc8330f	Bump upper bounds of dependencies	2020-01-18 21:40:47 +11:00
koenigmaximilian	04b026f5d7	Simplify recurrent example ... Produced better results when I tested (the Shakespeare example already contains a more complex recurrent network, let's KISS on this one.)	2018-07-07 20:57:25 +02:00
Huw Campbell	93435b2b99	Update mafia and make available in examples folder	2018-05-18 13:11:12 +10:00
Huw Campbell	9d9dfac6a2	And add a comment for where to grab the data	2018-01-04 16:27:37 +11:00
Huw Campbell	8349a85b0f	Give a slightly not terrible error message	2018-01-04 16:16:34 +11:00
Huw Campbell	0dd23bc2d2	Bump optparse in examples package	2017-12-15 09:13:30 +11:00
Huw Campbell	1ff89866be	Allow a recurrent net to be a recurrent layer	2017-07-14 14:05:41 +10:00
Huw Campbell	f9b53792fd	Network scale autodiff test	2017-04-11 10:08:08 +10:00
Huw Campbell	10a6003e73	Add Deconvolution layer and update the GAN example to use it ... This cuts the runtime by about 70% which is nice, and it's a better algorithm for it anyway. I've also refactored the Convolution layer such that there's only one actual implementation instead of two, and with that provided a few more instances for 2D and 3D shapes in and out. Update to the README and mnist show higher levels of composition.	2017-04-06 13:34:43 +10:00
Huw Campbell	e6293b8461	Cleanup imports and move examples to new project	2017-03-20 17:20:33 +11:00