def gen_backend(backend='cpu',
rng_seed=None,
default_dtype=np.float32,
batch_size=0,
stochastic_round=False,
device_id=0):
"""
Construct and return a backend instance of the appropriate type based on
the arguments given. With no parameters, a single CPU core, float32
backend is returned.
Arguments:
backend (string, optional): 'cpu' or 'gpu'.
rng_seed (numeric, optional): Set this to a numeric value which can be
used to seed the random number generator
of the instantiated backend. Defaults to
None, which doesn't explicitly seed (so
each run will be different)
default_dtype (dtype): Default tensor data type. CPU backend supports
np.float64, np.float32 and np.float16; GPU
backend supports np.float32 and np.float16.
batch_size (int): Set the size the data batches.
stochastic_round (int/bool, optional): Set this to True or an integer
to implent stochastic rounding.
If this is False rounding will
be to nearest.
If True will perform stochastic
rounding using default bit width.
If set to an integer will round
to that number of bits.
Only affects the gpu backend.
device_id (numeric, optional): Set this to a numeric value which can be
used to select which device to run the
process on
Returns:
Backend: newly constructed backend instance of the specifed type.
Notes:
* Attempts to construct a GPU instance without a CUDA capable card or
without nervanagpu package installed will cause the
program to display an error message and exit.
"""
logger = logging.getLogger(__name__)
if NervanaObject.be is not None:
# backend was already generated
# clean it up first
cleanup_backend()
else:
# at exit from python force cleanup of backend
# only register this function once, will use
# NervanaObject.be instead of a global
atexit.register(cleanup_backend)
if backend == 'cpu' or backend is None:
from neon.backends.nervanacpu import NervanaCPU
be = NervanaCPU(rng_seed=rng_seed, default_dtype=default_dtype)
elif backend == 'gpu':
gpuflag = False
# check nvcc
from neon.backends.util import check_gpu
gpuflag = (check_gpu.get_compute_capability(device_id) >= 5.0)
if gpuflag is False:
raise RuntimeError("Device " + str(device_id) +
" does not have CUDA compute " +
"capability 5.0 or greater")
from neon.backends.nervanagpu import NervanaGPU
# init gpu
be = NervanaGPU(rng_seed=rng_seed,
default_dtype=default_dtype,
stochastic_round=stochastic_round,
device_id=device_id)
elif backend == 'mgpu':
raise NotImplementedError("mgpu will be ready soon")
else:
raise ValueError("backend must be one of " "('cpu', 'gpu', 'mgpu')")
logger.info("Backend: {}, RNG seed: {}".format(backend, rng_seed))
NervanaObject.be = be
be.bsz = batch_size
return be
def gen_backend(backend='cpu', rng_seed=None, default_dtype=np.float32,
batch_size=0, stochastic_round=False, device_id=0):
"""
Construct and return a backend instance of the appropriate type based on
the arguments given. With no parameters, a single CPU core, float32
backend is returned.
Arguments:
backend (string, optional): 'cpu' or 'gpu'.
rng_seed (numeric, optional): Set this to a numeric value which can be
used to seed the random number generator
of the instantiated backend. Defaults to
None, which doesn't explicitly seed (so
each run will be different)
default_dtype (dtype): Default tensor data type. CPU backend supports
np.float64, np.float32 and np.float16; GPU
backend supports np.float32 and np.float16.
batch_size (int): Set the size the data batches.
stochastic_round (int/bool, optional): Set this to True or an integer
to implent stochastic rounding.
If this is False rounding will
be to nearest.
If True will perform stochastic
rounding using default bit width.
If set to an integer will round
to that number of bits.
Only affects the gpu backend.
device_id (numeric, optional): Set this to a numeric value which can be
used to select which device to run the
process on
Returns:
Backend: newly constructed backend instance of the specifed type.
Notes:
* Attempts to construct a GPU instance without a CUDA capable card or
without nervanagpu package installed will cause the
program to display an error message and exit.
"""
logger = logging.getLogger(__name__)
if NervanaObject.be is not None:
# backend was already generated
# clean it up first
cleanup_backend()
else:
# at exit from python force cleanup of backend
# only register this function once, will use
# NervanaObject.be instead of a global
atexit.register(cleanup_backend)
if backend == 'cpu' or backend is None:
from neon.backends.nervanacpu import NervanaCPU
be = NervanaCPU(rng_seed=rng_seed, default_dtype=default_dtype)
elif backend == 'gpu':
gpuflag = False
# check nvcc
from neon.backends.util import check_gpu
gpuflag = (check_gpu.get_compute_capability(device_id) >= 5.0)
if gpuflag is False:
raise RuntimeError("Device " + str(device_id) + " does not have CUDA compute " +
"capability 5.0 or greater")
from neon.backends.nervanagpu import NervanaGPU
# init gpu
be = NervanaGPU(rng_seed=rng_seed, default_dtype=default_dtype,
stochastic_round=stochastic_round, device_id=device_id)
elif backend == 'mgpu':
raise NotImplementedError("mgpu will be ready soon")
else:
raise ValueError("backend must be one of "
"('cpu', 'gpu', 'mgpu')")
logger.info("Backend: {}, RNG seed: {}".format(backend, rng_seed))
NervanaObject.be = be
be.bsz = batch_size
return be
def gen_backend(backend='cpu',
rng_seed=None,
datatype=np.float32,
batch_size=0,
stochastic_round=False,
device_id=0,
max_devices=get_device_count(),
compat_mode=None,
deterministic_update=False,
deterministic=True):
"""
Construct and return a backend instance of the appropriate type based on
the arguments given. With no parameters, a single CPU core, float32
backend is returned.
Arguments:
backend (string, optional): 'cpu' or 'gpu'.
rng_seed (numeric, optional): Set this to a numeric value which can be used to seed the
random number generator of the instantiated backend.
Defaults to None, which doesn't explicitly seed (so each run
will be different)
dataype (dtype): Default tensor data type. CPU backend supports np.float64, np.float32 and
np.float16; GPU backend supports np.float32 and np.float16.
batch_size (int): Set the size the data batches.
stochastic_round (int/bool, optional): Set this to True or an integer to implent
stochastic rounding. If this is False rounding will
be to nearest. If True will perform stochastic
rounding using default bit width. If set to an
integer will round to that number of bits.
Only affects the gpu backend.
device_id (numeric, optional): Set this to a numeric value which can be used to select
device on which to run the process
max_devices (int, optional): For use with multi-GPU backend only.
Controls the maximum number of GPUs to run
on.
compat_mode (str, optional): if this is set to 'caffe' then the conv and pooling
layer output sizes will match that of caffe as will
the dropout layer implementation
deterministic (bool, optional): if set to true, all operations will be done deterministically.
Returns:
Backend: newly constructed backend instance of the specifed type.
Notes:
* Attempts to construct a GPU instance without a CUDA capable card or without nervanagpu
package installed will cause the program to display an error message and exit.
"""
logger = logging.getLogger(__name__)
if NervanaObject.be is not None:
# backend was already generated clean it up first
cleanup_backend()
else:
# at exit from python force cleanup of backend only register this function once, will use
# NervanaObject.be instead of a global
atexit.register(cleanup_backend)
if deterministic_update:
deterministic = True
logger.warning(
"--deterministic_update is deprecated in favor of --deterministic")
if backend == 'cpu' or backend is None:
from neon.backends.nervanacpu import NervanaCPU
be = NervanaCPU(rng_seed=rng_seed,
default_dtype=datatype,
compat_mode=compat_mode)
elif backend == 'gpu' or backend == 'mgpu':
gpuflag = False
# check nvcc
from neon.backends.util import check_gpu
gpuflag = (check_gpu.get_compute_capability(device_id) >= 3.0)
if gpuflag is False:
raise RuntimeError("Device " + str(device_id) +
" does not have CUDA compute " +
"capability 3.0 or greater")
if backend == 'gpu':
from neon.backends.nervanagpu import NervanaGPU
# init gpu
be = NervanaGPU(rng_seed=rng_seed,
default_dtype=datatype,
stochastic_round=stochastic_round,
device_id=device_id,
compat_mode=compat_mode,
deterministic=deterministic)
else:
try:
from mgpu.nervanamgpu import NervanaMGPU
# init multiple GPU
be = NervanaMGPU(rng_seed=rng_seed,
default_dtype=datatype,
stochastic_round=stochastic_round,
num_devices=max_devices,
compat_mode=compat_mode,
deterministic=deterministic)
except ImportError:
logger.error(
"Multi-GPU support is a premium feature "
"available exclusively through the Nervana cloud."
" Please contact [email protected] for details.")
raise
elif backend == 'argon':
from argon.neon_backend.ar_backend import ArBackend
be = ArBackend(rng_seed=rng_seed, default_dtype=datatype)
else:
raise ValueError("backend must be one of ('cpu', 'gpu', 'mgpu')")
logger.info("Backend: {}, RNG seed: {}".format(backend, rng_seed))
NervanaObject.be = be
be.bsz = batch_size
return be
def gen_backend(backend='cpu', rng_seed=None, datatype=np.float32,
batch_size=0, stochastic_round=False, device_id=0,
max_devices=get_device_count(), compat_mode=None):
"""
Construct and return a backend instance of the appropriate type based on
the arguments given. With no parameters, a single CPU core, float32
backend is returned.
Arguments:
backend (string, optional): 'cpu' or 'gpu'.
rng_seed (numeric, optional): Set this to a numeric value which can be used to seed the
random number generator of the instantiated backend.
Defaults to None, which doesn't explicitly seed (so each run
will be different)
dataype (dtype): Default tensor data type. CPU backend supports np.float64, np.float32 and
np.float16; GPU backend supports np.float32 and np.float16.
batch_size (int): Set the size the data batches.
stochastic_round (int/bool, optional): Set this to True or an integer to implent
stochastic rounding. If this is False rounding will
be to nearest. If True will perform stochastic
rounding using default bit width. If set to an
integer will round to that number of bits.
Only affects the gpu backend.
device_id (numeric, optional): Set this to a numeric value which can be used to select
device on which to run the process
max_devices (int, optional): For use with multi-GPU backend only.
Controls the maximum number of GPUs to run
on.
compat_mode (str, optional): if this is set to 'caffe' then the conv and pooling
layer output sizes will match that of caffe as will
the dropout layer implementation
Returns:
Backend: newly constructed backend instance of the specifed type.
Notes:
* Attempts to construct a GPU instance without a CUDA capable card or without nervanagpu
package installed will cause the program to display an error message and exit.
"""
logger = logging.getLogger(__name__)
if NervanaObject.be is not None:
# backend was already generated clean it up first
cleanup_backend()
else:
# at exit from python force cleanup of backend only register this function once, will use
# NervanaObject.be instead of a global
atexit.register(cleanup_backend)
if backend == 'cpu' or backend is None:
from neon.backends.nervanacpu import NervanaCPU
be = NervanaCPU(rng_seed=rng_seed, default_dtype=datatype, compat_mode=compat_mode)
elif backend == 'gpu' or backend == 'mgpu':
gpuflag = False
# check nvcc
from neon.backends.util import check_gpu
gpuflag = (check_gpu.get_compute_capability(device_id) >= 5.0)
if gpuflag is False:
raise RuntimeError("Device " + str(device_id) + " does not have CUDA compute " +
"capability 5.0 or greater")
if backend == 'gpu':
from neon.backends.nervanagpu import NervanaGPU
# init gpu
be = NervanaGPU(rng_seed=rng_seed, default_dtype=datatype,
stochastic_round=stochastic_round,
device_id=device_id,
compat_mode=compat_mode)
else:
try:
from mgpu.nervanamgpu import NervanaMGPU
# init multiple GPU
be = NervanaMGPU(rng_seed=rng_seed,
default_dtype=datatype,
stochastic_round=stochastic_round,
num_devices=max_devices)
except ImportError:
logger.error("Multi-GPU support is a premium feature "
"available exclusively through the Nervana cloud."
" Please contact [email protected] for details.")
raise
else:
raise ValueError("backend must be one of ('cpu', 'gpu', 'mgpu')")
logger.info("Backend: {}, RNG seed: {}".format(backend, rng_seed))
NervanaObject.be = be
be.bsz = batch_size
return be
