• Github
Table of Contents
0.16.1.0
  • Welcome to PsyNeuLink
  • Basics and Primer
  • Quick Reference
  • Core
  • Library
  • Contributors Guide
  • Docs >
  • LLVM Compilation
Shortcuts

LLVM Compilation¶

class psyneulink.core.llvm.__init__.ExecutionMode(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)¶

Specify execution a Composition in interpreted or one of ithe compiled modes. These are used to specify the execution_mode argument of a Composition’s execute, run, and learn methods. See Compiled Modes under Compilation for additional details concerning use of each mode by a Composition.

Python¶

Execute using the Python interpreter; this is the default mode.

LLVM¶

compile and run Composition Nodes and Projections individually.

_LLVMExec¶

compile and run each TRIAL individually.

LLVMRun¶

compile and run multiple TRIALs.

Auto¶

progressively attempt LLVMRun, LLVMexec. LLVM and then Python.

PyTorch¶

execute the AutodiffComposition learn method using PyTorch, and its run method using the Python interpreter.

Warning

For clarity, this mode should only be used when executing an AutodiffComposition; using it with a standard Composition is possible, but it will not have the expected effect of executing its learn method using PyTorch.

PTX¶

compile and run Composition Nodes and Projections using CUDA for GPU.

PTXRun¶

compile and run multiple TRIALs using CUDA for GPU.


© Copyright 2016, Jonathan D. Cohen.

Built with Sphinx using a theme provided by Read the Docs.
  • LLVM Compilation
  • Github