[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: PDP++ v2.2 released!
Updates, especially of leabra which is subject to more tweaking and
adding of new features and removal of unused ones, typically result in
these messages -- they are almost always innocuous and disappear after
saving with the new version. However, it appears that somehow a
change in 2.2 has affected your project -- the main things that were
changed in 2.2 include:
- spike ->act_eq averaging (doesn't sound like you're using this)
- UNIT_GROUPS kwta computation -- this is most likely the problem --
see the ChangeLog comment under 2001-11-05 for details.
- TimeConSpec k_prv, k_frm were consolidated into a single INLINE
member, meaning that any settings of these params would be lost.
In general, you can find out most of these changes by looking in the
ChangeLog.. If these don't seem to be the problem, please send me
your old project that worked with 2.1 and I'll take a look -- you
might have found some kind of bug..
Hongbin Wang <Hongbin.Wang@uth.tmc.edu> writes:
> First of all, thanks a lot for the efforts and the great product.
>
> I built a leabra model early last year in version 2.0. After tuning, it
> fitted my behavioral data pretty well. When I tried to load the model in
> version 2.1 late last year, it gave (maaaany) the following messages:
> "*** Member: thr not found in type: LeabraUnit". Despite the warnings, the
> model was loaded successfully and was still able to run. And the fitting is
> no different.
> Now when I tried to load the same model in 2.2, the same (many) messages
> appeared plus many others, including a lot of:
> "*** Member: minus_delta_pt not found in type: LeabraLayerSpec
*** Member: acts_dif not found in type: LeabraLayer
*** Member: acts_dif not found in type: LeabraUnit_Group"
> Most importantly, though seems the model was finally loaded and displayed
> correctly, it didn't run correctly. For example, originally after 45 cycles
> the model would settle now it wouldn't until more than 200. So my fitting is
> now ruined.
> I wonder if you can point to a clue. Thanks a lot.
>
> Hongbin Wang
>
>
> > ----------
> > From: Randall C. O'Reilly
> > Reply To: oreilly@psych.colorado.edu
> > Sent: Thursday, January 17, 2002 11:57 PM
> > To: pdp-discuss@psych.colorado.edu
> > Subject: PDP++ v2.2 released!
> >
> > Version 2.2 is now available for downloading.. Note that the HP800
> > binaries are no longer supported, and that there were a few relatively
> > more serious bugs in the previous release that are now fixed..
> >
> > - Randy
> >
> > ANNOUNCING: The PDP++ Software, version 2.2
> >
> > Authors: Randall C. O'Reilly, Chadley K. Dawson, and James L. McClelland
> >
> > The PDP++ software is a neural-network simulation system written
> > in C++. It represents the next generation of the PDP software
> > released with the McClelland and Rumelhart "Explorations in Parallel
> > Distributed Processing Handbook", MIT Press, 1987. It is easy enough
> > for novice users, but very powerful and flexible for research use.
> >
> > The current version is 2.2, released January, 2002, which contains
> > some important new features, in addition to the usual bugfixes.
> >
> > The software can be obtained by anonymous ftp from:
> > Anonymous FTP Site: ftp://grey.colorado.edu/pub/oreilly/pdp++ *or*
> > ftp://cnbc.cmu.edu/pub/pdp++/ *or*
> > unix.hensa.ac.uk/mirrors/pdp++/
> >
> > For more information, see our web page:
> > WWW Page: http://www.cnbc.cmu.edu/PDP++/PDP++.html
> >
> > There is a 250 page (printed) manual and an HTML version available
> > on-line at the above address.
> >
> > The new features in 2.2 include:
> > ---------------------------------
> > o Fixed serious bugs with layer lesion flag, memory leak, OSX rnd #'s,
> > rbp real_time
> > o Auto-loading configuration files savable through gui
> > o Improved colorscale resolution
> > o Non-default value highlighting in edit dialogs
> > o Auto-scrolling windows during manipulation of nets, etc
> > o Many bug fixes and minor improvements
> >
> > See the file 'NEWS' for a more detailed summary of changes since
> > previous release.
> >
> > Software Features:
> > ==================
> >
> > o Full Graphical User Interface (GUI) based on the InterViews
> > toolkit. Allows user-selected "look and feel".
> >
> > o Network Viewer shows network architecture and processing in real-
> > time, allows network to be constructed with simple point-and-click
> > actions.
> >
> > o Training and testing data can be graphed on-line and network state
> > can be displayed over time numerically or using a wide range of
> > color or size-based graphical representations.
> >
> > o Environment Viewer shows training patterns using color or
> > size-based graphical representations; interactive configuration.
> >
> > o Flexible object-oriented design allows mix-and-match simulation
> > construction and easy extension by deriving new object types from
> > existing ones.
> >
> > o Built-in 'CSS' scripting language uses C++ syntax, allows full
> > access to simulation object data and functions. Transition
> > between script code and compiled code is simplified since both are
> > C++. Script has command-line completion, source-level debugger,
> > and provides standard C/C++ library functions and objects.
> >
> > o Scripts can control processing, generate training and testing
> > patterns, automate routine tasks, etc.
> >
> > o Scripts can be generated from GUI actions, and the user can create
> > GUI interfaces from script objects to extend and customize the
> > simulation environment.
> >
> >
> > Supported Algorithms:
> > =====================
> >
> > o Feedforward and recurrent error backpropagation. Recurrent BP
> > includes continuous, real-time models, and Almeida-Pineda.
> >
> > o Constraint satisfaction algorithms and associated learning
> > algorithms including Boltzmann Machine, Hopfield models,
> > mean-field networks (DBM), Interactive Activation and
> > Competition (IAC), and continuous stochastic networks.
> >
> > o Self-organizing learning including Competitive Learning, Soft
> > Competitive Learning, simple Hebbian, and Self-organizing Maps
> > ("Kohonen Nets").
> >
> > o Leabra algorithm that combines error-driven and Hebbian learning
> > with k-Winners-Take-All inhibitory competition. Over 40
> > research-grade simulations available for this algorithm in
> > association with new book: "Computational Explorations in
> > Cognitive Neuroscience: Understanding the Mind by Simulating the
> > Brain", O'Reilly & Munakata, 2000, MIT Press.
> >
> > The Fine Print:
> > ===============
> >
> > PDP++ is copyrighted and cannot be sold for profit by anyone other
> > than the copyright holders. However, the full source code is freely
> > available, and the user is granted full permission to modify, copy,
> > distribute and use it. See our web page for details.
> >
> > The software runs on Unix workstations under XWindows, and under MS
> > Windows (9X, NT, 2000). It requires a minimum of 16 Meg of RAM, and
> > 32 Meg is preferable. It has been developed and tested on Linux, Linux
> > PPC, Solaris, Irix, and HP-UX. Binaries are available for these
> > machines. Other machine types will require compiling from the source.
> >
> > The GUI in PDP++ is based on the InterViews toolkit, which we have
> > improved and maintained together with Michael Hines (developer of
> > NEURON). We distribute pre-compiled libraries for the supported
> > architectures, along with source code.
> >
> > The basic GUI and script technology in PDP++ is based on a
> > type-scanning system called TypeAccess which interfaces with the CSS
> > script language to provide a virtually automatic interface mechanism.
> > While these were developed for PDP++, they can easily be used for
> > any kind of application, and CSS is available as a stand-alone
> > executable for use like Perl or TCL.
> >
> > For more information on the details of the software, see our web page.
- Randy