Skip to content

Latest commit

 

History

History

code

Directory structure

abs_output location where results are saved

abstract.py logic for integration phase, some scaffolding for proposal phase

bb_rec_net.py back-bone of the recognition network

canon.py convert an inferred program into a standardized form

data.py represent primitive-based shapes, and how those shapes are explained by programs

dream.py logic for creating dreams used to train the recognition network

egg/Cargo.toml defines building logic for refactor opreation

egg/src/lib.rs defines egg refactor operation logic

env.yml conda environment to run code

extract.py run the refactor operation

library.py has library and language related logic

mask_info.py type-checking logic for recongition network during inference

mc_prog_search.py has logic for sampling from recognition network to explain input shapes

op_ord.py convert parametric expressions in abstractions into conditional rewrite logic

prm_expr_search.py search for parametric expressions to add to candidate abstraction functions

prm_sample_dist.py defines sampling distributions for parameters, used during the dreaming phase

prog_cost.py defines the objective function

prog_utils.py util functions related to programs

proposal.py main logic of proposal phase

rec_net.py training logic for recognition network

rewrites.py defines semantic rewrites for each domain

shapecoder.py entry-point for shapecoder method, connects the rest of the files

utils.py any other utility functions