Anatoly Belikov
Anatoly Belikov
My proposal is to add new flag to bindink function to reuse matched execution/evaluation links in pattern matcher search: Example: ```scheme (add-to-load-path ".") (use-modules (opencog)) (use-modules (opencog query)) (use-modules (opencog...
This issue is to document and discuss changes necessary to use deep learning frameworks with opencog. Our usecases: 1. We would like to be able to pass between ExecutionOutputLinks python...
Modified example from https://wiki.opencog.org/w/DualLink ```scheme (BindLink (ListLink (Concept "I") (Glob "$star") (Concept "you")) (ListLink (Concept "I") (Glob "$star") (Concept "you") (Concept "too"))) ``` ```scheme (cog-execute! (DualLink (ListLink (Concept "I") (VariableNode...
https://dev.singularitynet.io/tutorials/publish here step 4 is "Get ETH and AGI" but it doesn't describe where to get agi for main network. It is not clear where a developer should get agi...
Script to reproduce the bug: ``` from opencog.scheme import scheme_eval, scheme_eval_as from opencog.type_constructors import * atomspace = AtomSpace() set_default_atomspace(atomspace) scheme_eval(atomspace, '(use-modules (opencog))') #scheme_eval(atomspace, '(load "/usr/share/guile/site/2.2/opencog.scm")') scheme_eval(atomspace, '(TypeSet (Type "ConceptNode"))') #...
build_python_error_message and throw_on_error from PythonEval.cc looks quite useful for all code working with python c-api. Not sure if it should be moved to coguitls or just separate header in atomspace
works kinda better with very small descriptor length, though more experimentation is needed
lightweight notebook - working with preprocessed images
Following "Prepare questions dataset" section from experiments/opencog/question2atomese README I got file with invalid line ``` (InheritanceLink (ConceptNode """) (ConceptNode "size")) ``` Steps to reproduce: ``` wget http://visualqa.org/data/mscoco/vqa/v2_Questions_Train_mscoco.zip wget http://visualqa.org/data/mscoco/vqa/v2_Annotations_Train_mscoco.zip unzip...
# What does this PR do? logging loss for each image in dataset. Given that Lora datasets are usually small this feature might help to one to spot low-quality images...