Now everything is entirely handled through the command line
This commit is contained in:
Philippe Tillet
23
python/autotune/external/config.ini
vendored
23
python/autotune/external/config.ini
vendored
@@ -1,23 +0,0 @@
|
||||
#~ viennacl-src-root = /home/philippe/Development/viennacl-dev/viennacl/
|
||||
|
||||
[vector-axpy]
|
||||
precision = single
|
||||
#~ size = 5000000
|
||||
|
||||
#~ [reduction]
|
||||
#~ precision = single, double
|
||||
#~ size = 5000000
|
||||
#~
|
||||
#~ [matrix-axpy]
|
||||
#~ precision = single, double
|
||||
#~ size = 2560, 2560
|
||||
#~
|
||||
#~ [row-wise-reduction]
|
||||
#~ precision = single, double
|
||||
#~ layout = N,T
|
||||
#~ size = 2560, 2560
|
||||
#~
|
||||
#~ [matrix-product]
|
||||
#~ precision = single, double
|
||||
#~ layout = NN,NT,TN,TT
|
||||
#~ size = 1536, 1536, 1536
|
||||
@@ -14,9 +14,6 @@ from dataset import generate_dataset
|
||||
from model import train_model
|
||||
|
||||
|
||||
DATATYPES = { 'single' : vcl.float32,
|
||||
'double' : vcl.float64 }
|
||||
|
||||
TYPES = { 'vector-axpy': {'template':atd.VectorAxpyTemplate,
|
||||
'perf-index':lambda x: 3*x[0]*x[1][0]/x[2]*1e-9,
|
||||
'perf-measure':'GB/s'},
|
||||
@@ -38,125 +35,119 @@ TYPES = { 'vector-axpy': {'template':atd.VectorAxpyTemplate,
|
||||
'perf-measure': 'GFLOP/s'} }
|
||||
|
||||
|
||||
def do_tuning(config_fname, viennacl_root, device):
|
||||
def do_tuning(args, devices):
|
||||
json_out = {}
|
||||
config = ConfigObj(config_fname)
|
||||
device = devices[args.device]
|
||||
|
||||
def map_to_list(T, x):
|
||||
return list(map(T, x if isinstance(x, list) else [x]))
|
||||
|
||||
if(args.method=='unique'):
|
||||
default_tuning_sizes = {'vector-axpy': tuple(args.sizes[:1]), 'reduction': tuple(args.sizes[:1]),
|
||||
'matrix-axpy' : tuple(args.sizes[1:3]), 'row-wise-reduction' : tuple(args.sizes[1:3]),
|
||||
'matrix-product': tuple(args.sizes[3:])}
|
||||
for operation in ['vector-axpy', 'matrix-axpy', 'reduction', 'row-wise-reduction', 'matrix-product']:
|
||||
|
||||
if operation in config:
|
||||
p = config[operation]
|
||||
precisions = map_to_list(str, p['precision'])
|
||||
if 'all' in precisions:
|
||||
precisions = ['single','double']
|
||||
datatypes = [DATATYPES[k] for k in precisions]
|
||||
#Iterate through the datatypes
|
||||
for datatype in [vcl.float32, vcl.float64]:
|
||||
|
||||
#Iterate through the datatypes
|
||||
for datatype in datatypes:
|
||||
ctx = cl.Context([device])
|
||||
ctx = vcl.backend.Context(ctx)
|
||||
|
||||
ctx = cl.Context([device])
|
||||
ctx = vcl.backend.Context(ctx)
|
||||
#Check data-type
|
||||
if datatype is vcl.float64 and not device.double_fp_config:
|
||||
sys.stderr.write('Warning : The device ' + device.name + ' does not support double precision! Skipping ...')
|
||||
continue
|
||||
|
||||
#Check data-type
|
||||
if datatype is vcl.float64 and not device.double_fp_config:
|
||||
sys.stderr.write('Warning : The device ' + device.name + ' does not support double precision! Skipping ...')
|
||||
continue
|
||||
#Helper for execution
|
||||
def execute(device, node, other_params, sizes, fname = os.devnull, parameters = None):
|
||||
with vcl.Statement(node) as statement:
|
||||
if parameters:
|
||||
TemplateType = TYPES[operation]['template']
|
||||
return misc_tools.benchmark(TemplateType(TemplateType.Parameters(*parameters),*other_params), statement, device)
|
||||
print('-----')
|
||||
print(' '.join(map(str, ("Now tuning:", datatype.__name__, '-', operation, '-'.join(other_params), '[' + device.name, '(' + device.platform.name + ')] for sizes', sizes))))
|
||||
with open(fname, "w+") as archive:
|
||||
return optimize.genetic(statement, device, TYPES[operation]['template'], lambda p: TYPES[operation]['template'](p, *other_params),
|
||||
lambda t: TYPES[operation]['perf-index']([datatype().itemsize, sizes, t]), TYPES[operation]['perf-measure'], archive)
|
||||
|
||||
#Helper for execution
|
||||
def execute(device, node, other_params, sizes, fname = os.devnull, parameters = None):
|
||||
with vcl.Statement(node) as statement:
|
||||
if parameters:
|
||||
TemplateType = TYPES[operation]['template']
|
||||
return misc_tools.benchmark(TemplateType(TemplateType.Parameters(*parameters),*other_params), statement, device)
|
||||
print('-----')
|
||||
print(' '.join(map(str, ("Now tuning:", datatype.__name__, '-', operation, '-'.join(other_params), '[' + device.name, '(' + device.platform.name + ')] for sizes', sizes))))
|
||||
with open(fname, "w+") as archive:
|
||||
return optimize.genetic(statement, device, TYPES[operation]['template'], lambda p: TYPES[operation]['template'](p, *other_params),
|
||||
lambda t: TYPES[operation]['perf-index']([datatype().itemsize, sizes, t]), TYPES[operation]['perf-measure'], archive)
|
||||
#Helper for tuning
|
||||
def tune(execution_handler, nTuning, nDataPoints, draw, additional_parameters):
|
||||
#Update JSON
|
||||
full_operation = operation + ''.join(additional_parameters)
|
||||
if full_operation not in json_out:
|
||||
json_out[full_operation] = {}
|
||||
json_out[full_operation][datatype.__name__] = {}
|
||||
D = json_out[full_operation][datatype.__name__]
|
||||
|
||||
#Helper for tuning
|
||||
def tune(execution_handler, nTuning, nDataPoints, draw, additional_parameters):
|
||||
if 'size' in p:
|
||||
profile = execution_handler(map_to_list(int, p['size']))
|
||||
if 'viennacl-src-root' in config:
|
||||
misc_tools.update_viennacl_headers(config['viennacl-src-root'],device,datatype,operation,additional_parameters,profile)
|
||||
else:
|
||||
def compute_perf(x, t):
|
||||
return TYPES[operation]['perf-index']([datatype().itemsize, x, t])
|
||||
X, Y, profiles = generate_dataset(TYPES[operation]['template'], execution_handler, nTuning, nDataPoints, draw)
|
||||
clf = train_model(X, Y, profiles, TYPES[operation]['perf-measure'])
|
||||
if args.method == 'unique':
|
||||
profiles = [execution_handler(map(int,default_tuning_sizes[operation]))]
|
||||
if args.viennacl_src_path:
|
||||
misc_tools.update_viennacl_headers(args.viennacl_src_path,device,datatype,operation,additional_parameters,profiles[0])
|
||||
else:
|
||||
def compute_perf(x, t):
|
||||
return TYPES[operation]['perf-index']([datatype().itemsize, x, t])
|
||||
X, Y, profiles = generate_dataset(TYPES[operation]['template'], execution_handler, nTuning, nDataPoints, draw)
|
||||
clf = train_model(X, Y, profiles, TYPES[operation]['perf-measure'])
|
||||
D['predictor'] = [{'children_left': e.tree_.children_left.tolist(),
|
||||
'children_right': e.tree_.children_right.tolist(),
|
||||
'threshold': e.tree_.threshold.astype('float32').tolist(),
|
||||
'feature': e.tree_.feature.astype('float32').tolist(),
|
||||
'value': e.tree_.value[:,:,0].astype('float32').tolist()} for e in clf.estimators_]
|
||||
D['profiles'] = [ prof.astype('int').tolist() for prof in profiles]
|
||||
|
||||
#Update JSON
|
||||
full_operation = operation + ''.join(additional_parameters)
|
||||
if full_operation not in json_out:
|
||||
json_out[full_operation] = {}
|
||||
json_out[full_operation][datatype.__name__] = {}
|
||||
D = json_out[full_operation][datatype.__name__]
|
||||
D['profiles'] = [ prof.astype('int').tolist() for prof in profiles]
|
||||
D['predictor'] = [{'children_left': e.tree_.children_left.tolist(),
|
||||
'children_right': e.tree_.children_right.tolist(),
|
||||
'threshold': e.tree_.threshold.astype('float32').tolist(),
|
||||
'feature': e.tree_.feature.astype('float32').tolist(),
|
||||
'value': e.tree_.value[:,:,0].astype('float32').tolist()} for e in clf.estimators_]
|
||||
|
||||
#Vector AXPY
|
||||
if operation=='vector-axpy':
|
||||
def execution_handler(sizes, fname=os.devnull, parameters=None):
|
||||
x = vcl.Vector(sizes[0], context=ctx, dtype=datatype)
|
||||
y = vcl.Vector(sizes[0], context=ctx, dtype=datatype)
|
||||
z = vcl.Vector(sizes[0], context=ctx, dtype=datatype)
|
||||
return execute(device, vcl.Assign(z, vcl.ElementProd(vcl.exp(x + y),vcl.cos(x + y))), (), sizes, fname, parameters)
|
||||
tune(execution_handler, 30, 1000, lambda : 64*np.random.randint(low=10, high=100000, size=1), ())
|
||||
#Reduction
|
||||
if operation=='reduction':
|
||||
def execution_handler(sizes, fname=os.devnull, parameters=None):
|
||||
x = vcl.Vector(sizes[0], context=ctx, dtype=datatype)
|
||||
y = vcl.Vector(sizes[0], context=ctx, dtype=datatype)
|
||||
s = vcl.Scalar(0, context=ctx, dtype=datatype)
|
||||
return execute(device, vcl.Assign(s, vcl.Dot(x,y)), (), sizes, fname, parameters)
|
||||
tune(execution_handler, 50, 1000, lambda : 64*np.random.randint(low=10, high=100000, size=1), ())
|
||||
#Matrix AXPY
|
||||
if operation=='matrix-axpy':
|
||||
def execution_handler(sizes, fname=os.devnull, parameters=None):
|
||||
A = vcl.Matrix(sizes, context=ctx, dtype=datatype)
|
||||
B = vcl.Matrix(sizes, context=ctx, dtype=datatype)
|
||||
C = vcl.Matrix(sizes, context=ctx, dtype=datatype)
|
||||
return execute(device, vcl.Assign(C,A+B), (), sizes, fname, parameters)
|
||||
tune(execution_handler, 50, 1000, lambda : 64*np.random.randint(low=5, high=100, size=2), ())
|
||||
#Row-wise reduction
|
||||
if operation=='row-wise-reduction':
|
||||
layouts = map_to_list(str,p['layout'])
|
||||
if 'all' in layouts:
|
||||
layouts = ['N', 'T']
|
||||
for A_trans in layouts:
|
||||
def execution_handler(sizes, fname=os.devnull, parameters=None):
|
||||
A = vcl.Matrix(sizes if A_trans=='N' else sizes[::-1], context=ctx, dtype=datatype, layout=vcl.COL_MAJOR)
|
||||
x = vcl.Vector(sizes[1] if A_trans=='N' else sizes[0], context=ctx, dtype=datatype)
|
||||
y = vcl.Vector(sizes[0] if A_trans=='N' else sizes[1], context=ctx, dtype=datatype)
|
||||
LHS = A if A_trans=='N' else A.T
|
||||
return execute(device, vcl.Assign(y, LHS*x), (), sizes, fname, parameters)
|
||||
tune(execution_handler, 50, 1000, lambda : 64*np.random.randint(low=5, high=100, size=2), (A_trans,))
|
||||
#Matrix Product
|
||||
if operation=='matrix-product':
|
||||
layouts = map_to_list(str,p['layout'])
|
||||
if 'all' in layouts:
|
||||
layouts = ['NN', 'NT', 'TN', 'TT']
|
||||
for layout in layouts:
|
||||
def execution_handler(sizes, fname=os.devnull, parameters=None):
|
||||
A_trans = layout[0]
|
||||
B_trans = layout[1]
|
||||
A = vcl.Matrix((sizes[0], sizes[1]) if A_trans=='N' else (sizes[1],sizes[0]), context=ctx, dtype=datatype, layout=vcl.COL_MAJOR);
|
||||
B = vcl.Matrix((sizes[1], sizes[2]) if B_trans=='N' else (sizes[2],sizes[1]), context=ctx, dtype=datatype, layout=vcl.COL_MAJOR);
|
||||
LHS = A if A_trans=='N' else A.T
|
||||
RHS = B if B_trans=='N' else B.T
|
||||
alpha = vcl.HostScalar(1.0, context=ctx, dtype = datatype)
|
||||
beta = vcl.HostScalar(1.0, context=ctx, dtype = datatype)
|
||||
C = vcl.Matrix((sizes[0], sizes[2]), context=ctx, dtype = datatype, layout=vcl.COL_MAJOR)
|
||||
return execute(device, vcl.Assign(C,LHS*RHS*alpha + C*beta),(A_trans, B_trans), sizes, fname, parameters)
|
||||
tune(execution_handler, 50, 2000, lambda : 64*np.random.randint(low=1, high=40, size=3),(layout[0], layout[1]))
|
||||
#Vector AXPY
|
||||
if operation=='vector-axpy':
|
||||
def execution_handler(sizes, fname=os.devnull, parameters=None):
|
||||
x = vcl.Vector(sizes[0], context=ctx, dtype=datatype)
|
||||
y = vcl.Vector(sizes[0], context=ctx, dtype=datatype)
|
||||
z = vcl.Vector(sizes[0], context=ctx, dtype=datatype)
|
||||
return execute(device, vcl.Assign(z, vcl.ElementProd(vcl.exp(x + y),vcl.cos(x + y))), (), sizes, fname, parameters)
|
||||
tune(execution_handler, 30, 1000, lambda : 64*np.random.randint(low=10, high=100000, size=1), ())
|
||||
#Reduction
|
||||
if operation=='reduction':
|
||||
def execution_handler(sizes, fname=os.devnull, parameters=None):
|
||||
x = vcl.Vector(sizes[0], context=ctx, dtype=datatype)
|
||||
y = vcl.Vector(sizes[0], context=ctx, dtype=datatype)
|
||||
s = vcl.Scalar(0, context=ctx, dtype=datatype)
|
||||
return execute(device, vcl.Assign(s, vcl.Dot(x,y)), (), sizes, fname, parameters)
|
||||
tune(execution_handler, 30, 1000, lambda : 64*np.random.randint(low=10, high=100000, size=1), ())
|
||||
#Matrix AXPY
|
||||
if operation=='matrix-axpy':
|
||||
def execution_handler(sizes, fname=os.devnull, parameters=None):
|
||||
A = vcl.Matrix(sizes, context=ctx, dtype=datatype)
|
||||
B = vcl.Matrix(sizes, context=ctx, dtype=datatype)
|
||||
C = vcl.Matrix(sizes, context=ctx, dtype=datatype)
|
||||
return execute(device, vcl.Assign(C,A+B), (), sizes, fname, parameters)
|
||||
tune(execution_handler, 30, 1000, lambda : 64*np.random.randint(low=5, high=100, size=2), ())
|
||||
#Row-wise reduction
|
||||
if operation=='row-wise-reduction':
|
||||
layouts = ['N', 'T']
|
||||
for A_trans in layouts:
|
||||
def execution_handler(sizes, fname=os.devnull, parameters=None):
|
||||
A = vcl.Matrix(sizes if A_trans=='N' else sizes[::-1], context=ctx, dtype=datatype, layout=vcl.COL_MAJOR)
|
||||
x = vcl.Vector(sizes[1] if A_trans=='N' else sizes[0], context=ctx, dtype=datatype)
|
||||
y = vcl.Vector(sizes[0] if A_trans=='N' else sizes[1], context=ctx, dtype=datatype)
|
||||
LHS = A if A_trans=='N' else A.T
|
||||
return execute(device, vcl.Assign(y, LHS*x), (), sizes, fname, parameters)
|
||||
tune(execution_handler, 30, 1000, lambda : 64*np.random.randint(low=5, high=100, size=2), (A_trans,))
|
||||
#Matrix Product
|
||||
if operation=='matrix-product':
|
||||
layouts = ['NN', 'NT', 'TN', 'TT']
|
||||
for layout in layouts:
|
||||
def execution_handler(sizes, fname=os.devnull, parameters=None):
|
||||
A_trans = layout[0]
|
||||
B_trans = layout[1]
|
||||
A = vcl.Matrix((sizes[0], sizes[1]) if A_trans=='N' else (sizes[1],sizes[0]), context=ctx, dtype=datatype, layout=vcl.COL_MAJOR);
|
||||
B = vcl.Matrix((sizes[1], sizes[2]) if B_trans=='N' else (sizes[2],sizes[1]), context=ctx, dtype=datatype, layout=vcl.COL_MAJOR);
|
||||
LHS = A if A_trans=='N' else A.T
|
||||
RHS = B if B_trans=='N' else B.T
|
||||
alpha = vcl.HostScalar(1.0, context=ctx, dtype = datatype)
|
||||
beta = vcl.HostScalar(1.0, context=ctx, dtype = datatype)
|
||||
C = vcl.Matrix((sizes[0], sizes[2]), context=ctx, dtype = datatype, layout=vcl.COL_MAJOR)
|
||||
return execute(device, vcl.Assign(C,LHS*RHS*alpha + C*beta),(A_trans, B_trans), sizes, fname, parameters)
|
||||
tune(execution_handler, 30, 1000, lambda : 64*np.random.randint(low=1, high=40, size=3),(layout[0], layout[1]))
|
||||
|
||||
dname = misc_tools.sanitize_string(device.name)
|
||||
json_out["version"] = "1.0"
|
||||
@@ -169,9 +160,16 @@ if __name__ == "__main__":
|
||||
subparsers = parser.add_subparsers(dest='action')
|
||||
print_devices_parser = subparsers.add_parser('list-devices', help='list the devices available')
|
||||
tune_parser = subparsers.add_parser('tune', help='tune using a specific configuration file')
|
||||
tune_parser.add_argument("--config", default="config.ini", required=False, type=str)
|
||||
tune_parser.add_argument("--device", default=0, required=False, type=str)
|
||||
tune_parser.add_argument("--viennacl-root", default='', required=False, type=str)
|
||||
|
||||
tune_subparsers = tune_parser.add_subparsers(dest='method')
|
||||
big_sizes_parser = tune_subparsers.add_parser('unique', help = 'Tune each operation for unique sizes')
|
||||
big_sizes_parser.add_argument("--sizes", nargs='+', default=[10e6,2560,2560,1536,1536,1536], required=False, type=int, help = '6 = 1 + 2 + 3 sizes for respectively BLAS1, BLAS2, BLAS3')
|
||||
big_sizes_parser.add_argument("--viennacl-src-path", default='', required=False, type=str)
|
||||
|
||||
model_parser = tune_subparsers.add_parser('build-model', help = 'Build an input-dependent model')
|
||||
|
||||
|
||||
args = parser.parse_args()
|
||||
|
||||
devices = [d for platform in cl.get_platforms() for d in platform.get_devices()]
|
||||
@@ -186,4 +184,4 @@ if __name__ == "__main__":
|
||||
print("------")
|
||||
print("Auto-tuning")
|
||||
print("------")
|
||||
do_tuning(args.config, args.viennacl_root, devices[args.device])
|
||||
do_tuning(args, devices)
|
||||
|
||||
Reference in New Issue
Block a user