def init_backend_engine():
"""
Initializes ``engine``, which is either :class:`TFEngine.Engine` or Theano :class:`Engine.Engine`.
"""
BackendEngine.select_engine(config=config)
if BackendEngine.is_theano_selected():
print("Theano:", describe_theano_version(), file=log.v3)
import TheanoUtil
TheanoUtil.monkey_patches()
elif BackendEngine.is_tensorflow_selected():
print("TensorFlow:", describe_tensorflow_version(), file=log.v3)
if get_tensorflow_version_tuple()[0] == 0:
print("Warning: TF <1.0 is not supported and likely broken.", file=log.v2)
if os.environ.get("TF_DEVICE"):
print("Devices: Use %s via TF_DEVICE instead of %s." % (
os.environ.get("TF_DEVICE"), config.opt_typed_value("device")), file=log.v4)
config.set("device", os.environ.get("TF_DEVICE"))
if config.is_true("use_horovod"):
import socket
# noinspection PyPackageRequirements,PyUnresolvedReferences
import horovod.tensorflow as hvd
from TFUtil import init_horovod
init_horovod() # make sure it is initialized
if "gpu" in config.value("device", "") or os.environ.get("CUDA_VISIBLE_DEVICES", ""):
# We assume that we want to use a GPU.
gpu_opts = config.typed_dict.setdefault("tf_session_opts", {}).setdefault("gpu_options", {})
assert "visible_device_list" not in gpu_opts
gpu_opts["visible_device_list"] = str(hvd.local_rank())
print("Horovod: Hostname %s, pid %i, using GPU %s." % (
socket.gethostname(), os.getpid(), gpu_opts["visible_device_list"]), file=log.v3)
else:
if hvd.rank() == 0: # Don't spam in all ranks.
print("Horovod: Not using GPU.", file=log.v3)
horovod_reduce_type = config.value("horovod_reduce_type", "")
if horovod_reduce_type == "":
horovod_reduce_type = "grad"
config.set("horovod_reduce_type", horovod_reduce_type)
else:
assert horovod_reduce_type in ["grad", "param"], "config option 'horovod_reduce_type' invalid"
if hvd.rank() == 0: # Don't spam in all ranks.
print("Horovod: Reduce type:", horovod_reduce_type, file=log.v3)
from TFUtil import debug_register_better_repr, setup_tf_thread_pools, print_available_devices
tf_session_opts = config.typed_value("tf_session_opts", {})
assert isinstance(tf_session_opts, dict)
# This must be done after the Horovod logic, such that we only touch the devices we are supposed to touch.
setup_tf_thread_pools(log_file=log.v3, tf_session_opts=tf_session_opts)
# Print available devices. Also make sure that get_tf_list_local_devices uses the correct TF session opts.
print_available_devices(tf_session_opts=tf_session_opts, file=log.v2)
debug_register_better_repr()
else:
raise NotImplementedError
def init_backend_engine():
"""
Initializes ``engine``, which is either :class:`TFEngine.Engine` or Theano :class:`Engine.Engine`.
"""
BackendEngine.select_engine(config=config)
if BackendEngine.is_theano_selected():
print("Theano:", describe_theano_version(), file=log.v3)
import TheanoUtil
TheanoUtil.monkey_patches()
elif BackendEngine.is_tensorflow_selected():
print("TensorFlow:", describe_tensorflow_version(), file=log.v3)
if get_tensorflow_version_tuple()[0] == 0:
print("Warning: TF <1.0 is not supported and likely broken.", file=log.v2)
if os.environ.get("TF_DEVICE"):
print("Devices: Use %s via TF_DEVICE instead of %s." % (
os.environ.get("TF_DEVICE"), config.opt_typed_value("device")), file=log.v4)
config.set("device", os.environ.get("TF_DEVICE"))
if config.is_true("use_horovod"):
import socket
# noinspection PyPackageRequirements,PyUnresolvedReferences
import horovod.tensorflow as hvd
from TFUtil import init_horovod
init_horovod() # make sure it is initialized
if "gpu" in config.value("device", "") or os.environ.get("CUDA_VISIBLE_DEVICES", ""):
# We assume that we want to use a GPU.
gpu_opts = config.typed_dict.setdefault("tf_session_opts", {}).setdefault("gpu_options", {})
assert "visible_device_list" not in gpu_opts
gpu_opts["visible_device_list"] = str(hvd.local_rank())
print("Horovod: Hostname %s, pid %i, using GPU %s." % (
socket.gethostname(), os.getpid(), gpu_opts["visible_device_list"]), file=log.v3)
else:
if hvd.rank() == 0: # Don't spam in all ranks.
print("Horovod: Not using GPU.", file=log.v3)
horovod_reduce_type = config.value("horovod_reduce_type", "")
if horovod_reduce_type == "":
horovod_reduce_type = "grad"
config.set("horovod_reduce_type", horovod_reduce_type)
else:
assert horovod_reduce_type in ["grad", "param"], "config option 'horovod_reduce_type' invalid"
if hvd.rank() == 0: # Don't spam in all ranks.
print("Horovod: Reduce type:", horovod_reduce_type, file=log.v3)
from TFUtil import debug_register_better_repr, setup_tf_thread_pools, print_available_devices
tf_session_opts = config.typed_value("tf_session_opts", {})
assert isinstance(tf_session_opts, dict)
# This must be done after the Horovod logic, such that we only touch the devices we are supposed to touch.
setup_tf_thread_pools(log_file=log.v3, tf_session_opts=tf_session_opts)
# Print available devices. Also make sure that get_tf_list_local_devices uses the correct TF session opts.
print_available_devices(tf_session_opts=tf_session_opts, file=log.v2)
debug_register_better_repr()
else:
raise NotImplementedError