#!/usr/bin/python
# -*- coding: utf-8 -*-
##
# gpu_models.py: Demonstrates the use of GPU-accelerated likelihood evaluation.
##
# © 2014 Chris Ferrie (csferrie@gmail.com) and
# Christopher E. Granade (cgranade@gmail.com).
#
# Licensed under the AGPL version 3.
##
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU Affero General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU Affero General Public License for more details.
#
# You should have received a copy of the GNU Affero General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
##
"""
This module demonstrates the use of OpenCL to accelerate sequential Monte Carlo
by implementing a simple cosine-likelihood model as an OpenCL kernel. When run
as a script, this module then compares the performance of the OpenCL-accelerated
model to the pure-NumPy model implemented in the QInfer project.
"""
## FEATURES ####################################################################
from __future__ import absolute_import
from __future__ import print_function
from __future__ import division
## EXPORTS #####################################################################
__all__ = [
'AcceleratedPrecessionModel'
]
## IMPORTS #####################################################################
from builtins import range
try:
import pyopencl as cl
except ImportError:
cl = None
import warnings
warnings.warn(
"Could not import PyOpenCL. GPU models will not work.",
ImportWarning
)
import numpy as np
import numpy.linalg as la
import time
from qinfer.abstract_model import Model, FiniteOutcomeModel
from qinfer.test_models import SimplePrecessionModel
from qinfer.smc import SMCUpdater
from qinfer.distributions import UniformDistribution
## KERNELS #####################################################################
COS_MODEL_KERNEL = """
__kernel void cos_model(
int n_experiments,
__global const float *models,
__global const float *expparams,
__global float *likelihoods
) {
// Assuming two-outcome model, and finding Pr(0 | model; expparams).
int idx_model = get_global_id(0);
int idx_experiment = get_global_id(1);
likelihoods[idx_model * n_experiments + idx_experiment] = pow(cos(
models[idx_model] * expparams[idx_experiment] / 2
), 2);
}
"""
## CLASSES #####################################################################
[docs]class AcceleratedPrecessionModel(FiniteOutcomeModel):
r"""
Reimplementation of `qinfer.test_models.SimplePrecessionModel`, using OpenCL
to accelerate computation.
"""
def __init__(self, context=None):
super(AcceleratedPrecessionModel, self).__init__()
if cl is None:
raise ImportError(
"AcceleratedPrecessionModel requires "
"GPU acceleration support. Please install PyOpenCL to "
"use this model."
)
self._ctx = cl.create_some_context() if context is None else context
self._queue = cl.CommandQueue(self._ctx)
self._prg = cl.Program(self._ctx, COS_MODEL_KERNEL).build()
## PROPERTIES ##
@property
def n_modelparams(self):
return 1
@property
def expparams_dtype(self):
return 'float'
@property
def is_n_outcomes_constant(self):
"""
Returns ``True`` if and only if the number of outcomes for each
experiment is independent of the experiment being performed.
This property is assumed by inference engines to be constant for
the lifetime of a Model instance.
"""
return True
## METHODS ##
@staticmethod
[docs] def are_models_valid(modelparams):
return np.all(
modelparams > 0,
axis=1
)
[docs] def n_outcomes(self, expparams):
"""
Returns an array of dtype ``uint`` describing the number of outcomes
for each experiment specified by ``expparams``.
:param numpy.ndarray expparams: Array of experimental parameters. This
array must be of dtype agreeing with the ``expparams_dtype``
property.
"""
return 2
[docs] def likelihood(self, outcomes, modelparams, expparams):
# By calling the superclass implementation, we can consolidate
# call counting there.
super(AcceleratedPrecessionModel, self).likelihood(outcomes, modelparams, expparams)
# Possibly add a second axis to modelparams.
if len(modelparams.shape) == 1:
modelparams = modelparams[..., np.newaxis]
# Convert to float32 if needed.
mps = modelparams.astype(np.float32)
eps = expparams.astype(np.float32)
# Allocating a buffer for the pr0 returns.
pr0 = np.empty((mps.shape[0], eps.shape[0]), dtype=mps.dtype)
# Move buffers to the GPU.
mf = cl.mem_flags
mps_buf = cl.Buffer(self._ctx, mf.READ_ONLY | mf.COPY_HOST_PTR, hostbuf=mps)
eps_buf = cl.Buffer(self._ctx, mf.READ_ONLY | mf.COPY_HOST_PTR, hostbuf=eps)
dest_buf = cl.Buffer(self._ctx, mf.WRITE_ONLY, pr0.nbytes)
# Run the kernel with global worksize (n_models, n_experiments).
self._prg.cos_model(self._queue, pr0.shape, None, np.int32(eps.shape[0]), mps_buf, eps_buf, dest_buf)
# Copy the buffer back from the GPU and free memory there.
cl.enqueue_copy(self._queue, pr0, dest_buf)
mps_buf.release()
eps_buf.release()
dest_buf.release()
# Now we concatenate over outcomes.
return FiniteOutcomeModel.pr0_to_likelihood_array(outcomes, pr0)
## SCRIPT ######################################################################
if __name__ == "__main__":
# NOTE: This is now redundant with the perf_testing module.
simple_model = SimplePrecessionModel()
for model in [AcceleratedPrecessionModel(), SimplePrecessionModel()]:
true = np.random.random(1)
updater = SMCUpdater(model, 100000, UniformDistribution([0, 1]))
tic = time.time()
for idx_exp in range(200):
if not (idx_exp % 20):
print(idx_exp)
expparams = np.array([(9 / 8) ** idx_exp])
updater.update(simple_model.simulate_experiment(true, expparams), expparams)
print(model, updater.est_mean(), true, time.time() - tic)