Leabra problem with updated installtion
Stephen Read
After tinkering around I realized that for some reason go mod init and go mod tidy, were setting up a version of emergent of 1.4.21, which seems to be flagged on github as a failure. When I changed emergent’s version to 1.3.52 all the errors below disappeared, but now I’m getting the following error.
# Depress
./logitems.go:115:51: cannot convert 0 (untyped int constant) to []float64
./logitems.go:118:18: cannot convert 0 (untyped int constant) to []float64
As I wrote in an earlier email, at least three of my students were getting the following message when they created their own code directory, using ra25. I just updated the go.mod in one of my code directories for another program because something I built a year ago was now crashing. When I did that I now get the following errors. I have also tried with lea...@v1.2.5 and 1.2.9 and get the identical error, except that the version number is different.
One of my students spent hours on this and finally got it to work, but he has no idea how.
When I originally built my program it was using lea...@v1.2.7, so I don’t understand what is suddenly going on. I’m running Sonoma 14.2.1.
# github.com/emer/leabra/leabra
../../go/pkg/mod/github.com/emer/lea...@v1.2.7/leabra/layer.go:155:45: not enough arguments in call to ly.LeabraLay.UnitVal1D
have (int, int)
want (int, int, int)
../../go/pkg/mod/github.com/emer/lea...@v1.2.7/leabra/layer.go:178:37: not enough arguments in call to ly.LeabraLay.UnitVal1D
have (int, int)
want (int, int, int)
../../go/pkg/mod/github.com/emer/lea...@v1.2.7/leabra/layer.go:220:37: not enough arguments in call to ly.LeabraLay.UnitVal1D
have (int, int)
want (int, int, int)
../../go/pkg/mod/github.com/emer/lea...@v1.2.7/leabra/layer.go:238:38: not enough arguments in call to ly.LeabraLay.UnitVal1D
have (int, int)
want (int, int, int)
../../go/pkg/mod/github.com/emer/lea...@v1.2.7/leabra/layer.go:348:26: cannot use ly (variable of type *Layer) as type emer.Layer in argument to emer.SendNameTry:
*Layer does not implement emer.Layer (missing AddClass method)
../../go/pkg/mod/github.com/emer/lea...@v1.2.7/leabra/layer.go:351:30: cannot use ly (variable of type *Layer) as type emer.Layer in argument to emer.SendNameTypeTry:
*Layer does not implement emer.Layer (missing AddClass method)
../../go/pkg/mod/github.com/emer/lea...@v1.2.7/leabra/layer.go:354:26: cannot use ly (variable of type *Layer) as type emer.Layer in argument to emer.RecvNameTry:
*Layer does not implement emer.Layer (missing AddClass method)
../../go/pkg/mod/github.com/emer/lea...@v1.2.7/leabra/layer.go:357:30: cannot use ly (variable of type *Layer) as type emer.Layer in argument to emer.RecvNameTypeTry:
*Layer does not implement emer.Layer (missing AddClass method)
../../go/pkg/mod/github.com/emer/lea...@v1.2.7/leabra/network.go:39:9: cannot use &Layer{} (value of type *Layer) as type emer.Layer in return statement:
*Layer does not implement emer.Layer (missing AddClass method)
../../go/pkg/mod/github.com/emer/lea...@v1.2.7/leabra/network.go:44:9: cannot use &Prjn{} (value of type *Prjn) as type emer.Prjn in return statement:
*Prjn does not implement emer.Prjn (missing AddClass method)
../../go/pkg/mod/github.com/emer/lea...@v1.2.7/leabra/network.go:44:9: too many errors
Stephen J. Read
Mendel B. Silberberg Professor of Social Psychology
Department of Psychology
University of Southern California
Los Angeles, CA 90089-1061
Website: www.stephenjread.com
Randall O'Reilly
I just pushed updated go.mod for leabra to these versions, which are the latest ones in the v1 branches:
require (
github.com/c2h5oh/datasize v0.0.0-20220606134207-859f65c6625b
github.com/emer/emergent v1.3.53
github.com/emer/empi v1.0.19
github.com/emer/etable v1.1.24
github.com/goki/gi v1.3.25
github.com/goki/ki v1.1.17
github.com/goki/kigen v1.0.2
github.com/goki/mat32 v1.0.18
)
for me, this builds all the examples and I spot checked a few and all were working as expected.
I can see in examples/ra25/logitems.go, that there was a fix for that exact build issue you're having, associated with a change in the emergent code, in going from 1.3.50 -> 1.3.52:
https://github.com/emer/leabra/commit/c05e2dc28d06276fce467b7de5272bd8013d4988
should now be:
if ss.Stats.Int("FirstZero") < 0 && pcterr[0] == 0 {
was before: pcterr == 0 at the end there -- it is now a slice.
Hope that helps,
- Randy
> You received this message because you are subscribed to the Google Groups "emergent-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to emergent-user...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/emergent-users/BYAPR07MB4679F956CCB4B4E02156F9F2D7502%40BYAPR07MB4679.namprd07.prod.outlook.com.
Randall O'Reilly
Begin forwarded message:From: Randall O'Reilly <rcore...@gmail.com>Subject: Re: Leabra problem with updated installtionDate: March 7, 2024 at 11:59:22 AM PSTTo: Stephen Read <re...@usc.edu>That is fixed by the more recent versions of gi and vgpu -- you can independently just update gi to 1.3.25 to fix that for example.
- RandyOn Mar 7, 2024, at 11:52 AM, Stephen Read <re...@usc.edu> wrote:
<!-- /* Font Definitions */ @font-face {font-family:"Cambria Math"; panose-1:2 4 5 3 5 4 6 3 2 4;} @font-face {font-family:Calibri; panose-1:2 15 5 2 2 2 4 3 2 4;} @font-face {font-family:Aptos; panose-1:2 11 0 4 2 2 2 2 2 4;} @font-face {font-family:"Times New Roman \(Body CS\)"; panose-1:2 11 6 4 2 2 2 2 2 4;} /* Style Definitions */ p.MsoNormal, li.MsoNormal, div.MsoNormal {margin:0in; font-size:10.0pt; font-family:"Aptos",sans-serif;} a:link, span.MsoHyperlink {mso-style-priority:99; color:blue; text-decoration:underline;} span.EmailStyle19 {mso-style-type:personal-reply; font-family:"Aptos",sans-serif; color:windowtext;} .MsoChpDefault {mso-style-type:export-only; font-size:10.0pt; mso-ligatures:none;} @page WordSection1 {size:8.5in 11.0in; margin:1.0in 1.0in 1.0in 1.0in;} div.WordSection1 {page:WordSection1;} --> Thank you for doing this so quickly.
One other thing. For some reason at least 4 or 5 of my students have gotten a message about having a missing rpath, and not being able to find the dylib files for the Mac in the VulkanSDK. They have figured out how to fix it, but I have never seen this error before. Here is a png of one student’s terminal.<image001.png> Stephen J. Read
Mendel B. Silberberg Professor of Social Psychology
Department of Psychology
University of Southern California
Los Angeles, CA 90089-1061
Website: www.stephenjread.com
From: Randall O'Reilly <rcore...@gmail.com>
Date: Thursday, March 7, 2024 at 11:38 AM
To: Stephen Read <re...@usc.edu>
Cc: emergen...@googlegroups.com <emergen...@googlegroups.com>
Subject: Re: Leabra problem with updated installtion
Steve,
I just pushed updated go.mod for leabra to these versions, which are the latest ones in the v1 branches:
require (
github.com/c2h5oh/datasize v0.0.0-20220606134207-859f65c6625b
github.com/emer/emergent v1.3.53
github.com/emer/empi v1.0.19
github.com/emer/etable v1.1.24
github.com/goki/gi v1.3.25
github.com/goki/ki v1.1.17
github.com/goki/kigen v1.0.2
github.com/goki/mat32 v1.0.18
)
for me, this builds all the examples and I spot checked a few and all were working as expected.
I can see in examples/ra25/logitems.go, that there was a fix for that exact build issue you're having, associated with a change in the emergent code, in going from 1.3.50 -> 1.3.52:
--
You received this message because you are subscribed to the Google Groups "emergent-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to emergent-user...@googlegroups.com.
To view this discussion on the web visit https://urldefense.com/v3/__https://groups.google.com/d/msgid/emergent-users/BYAPR07MB4679F956CCB4B4E02156F9F2D7502*40BYAPR07MB4679.namprd07.prod.outlook.com__;JQ!!LIr3w8kk_Xxm!ptMiSUsen4chH22ahk0WcEH8duVmQn1Q8GosZ7XqOz9hQDTbw15id_PapI0w0SDKwtFMWB_3Zh_-B1U$ .
Stephen Read
Thank you again.
To view this discussion on the web visit https://groups.google.com/d/msgid/emergent-users/71CF457E-9949-4307-8D72-86985E9D2FEB%40gmail.com.