Bagua API Documentation
v0.8.2
API Documents
bagua
Bagua API Documentation
Bagua API Documentation
»
Index
Edit on GitHub
Index
A
|
B
|
C
|
D
|
E
|
F
|
G
|
I
|
L
|
M
|
N
|
P
|
Q
|
R
|
S
|
T
|
U
|
V
|
W
A
abort() (bagua.torch_api.algorithms.async_model_average.AsyncModelAverageAlgorithmImpl method)
Algorithm (class in bagua.torch_api.algorithms)
(class in bagua.torch_api.algorithms.base)
AlgorithmImpl (class in bagua.torch_api.algorithms)
(class in bagua.torch_api.algorithms.base)
allgather() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
allgather_inplace() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
allreduce() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
allreduce_inplace() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
alltoall() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
alltoall_inplace() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
alltoall_v() (in module bagua.torch_api)
alltoall_v_inplace() (in module bagua.torch_api)
append_asynchronous_model_average_op() (bagua.torch_api.bucket.BaguaBucket method)
append_centralized_synchronous_op() (bagua.torch_api.bucket.BaguaBucket method)
append_decentralized_synchronous_op() (bagua.torch_api.bucket.BaguaBucket method)
append_low_precision_decentralized_synchronous_op() (bagua.torch_api.bucket.BaguaBucket method)
append_python_op() (bagua.torch_api.bucket.BaguaBucket method)
AsyncModelAverageAlgorithm (class in bagua.torch_api.algorithms.async_model_average)
AsyncModelAverageAlgorithmImpl (class in bagua.torch_api.algorithms.async_model_average)
AVG (bagua.torch_api.communication.ReduceOp attribute)
(bagua.torch_api.ReduceOp attribute)
B
bagua
module
bagua.torch_api
module
bagua.torch_api.algorithms
module
bagua.torch_api.algorithms.async_model_average
module
bagua.torch_api.algorithms.base
module
bagua.torch_api.algorithms.bytegrad
module
bagua.torch_api.algorithms.decentralized
module
bagua.torch_api.algorithms.gradient_allreduce
module
bagua.torch_api.algorithms.q_adam
module
bagua.torch_api.bucket
module
bagua.torch_api.checkpoint
module
bagua.torch_api.checkpoint.checkpointing
module
bagua.torch_api.communication
module
bagua.torch_api.contrib
module
bagua.torch_api.contrib.cache_loader
module
bagua.torch_api.contrib.cached_dataset
module
bagua.torch_api.contrib.fuse
module
bagua.torch_api.contrib.fuse.optimizer
module
bagua.torch_api.contrib.load_balancing_data_loader
module
bagua.torch_api.contrib.sync_batchnorm
module
bagua.torch_api.contrib.utils
module
bagua.torch_api.contrib.utils.redis_store
module
bagua.torch_api.contrib.utils.store
module
bagua.torch_api.distributed
module
bagua.torch_api.env
module
bagua.torch_api.model_parallel
module
bagua.torch_api.model_parallel.moe
module
bagua.torch_api.model_parallel.moe.experts
module
bagua.torch_api.model_parallel.moe.layer
module
bagua.torch_api.model_parallel.moe.sharded_moe
module
bagua.torch_api.model_parallel.moe.utils
module
bagua.torch_api.tensor
module
bagua_backend_tensor() (bagua.torch_api.BaguaTensor method)
(bagua.torch_api.tensor.BaguaTensor method)
bagua_build_params() (bagua.torch_api.BaguaModule method)
(bagua.torch_api.distributed.BaguaModule method)
bagua_ensure_grad() (bagua.torch_api.BaguaTensor method)
(bagua.torch_api.tensor.BaguaTensor method)
bagua_getter_closure() (bagua.torch_api.BaguaTensor method)
(bagua.torch_api.tensor.BaguaTensor method)
bagua_mark_communication_ready() (bagua.torch_api.BaguaTensor method)
(bagua.torch_api.tensor.BaguaTensor method)
bagua_mark_communication_ready_without_synchronization() (bagua.torch_api.BaguaTensor method)
(bagua.torch_api.tensor.BaguaTensor method)
bagua_set_storage() (bagua.torch_api.BaguaTensor method)
(bagua.torch_api.tensor.BaguaTensor method)
bagua_setter_closure() (bagua.torch_api.BaguaTensor method)
(bagua.torch_api.tensor.BaguaTensor method)
BaguaBucket (class in bagua.torch_api.bucket)
BaguaModule (class in bagua.torch_api)
(class in bagua.torch_api.distributed)
BaguaProcessGroup (class in bagua.torch_api.communication)
BaguaTensor (class in bagua.torch_api)
(class in bagua.torch_api.tensor)
BAND (bagua.torch_api.communication.ReduceOp attribute)
(bagua.torch_api.ReduceOp attribute)
barrier() (in module bagua.torch_api.communication)
Base (in module bagua.torch_api.model_parallel.moe.sharded_moe)
BOR (bagua.torch_api.communication.ReduceOp attribute)
(bagua.torch_api.ReduceOp attribute)
broadcast() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
BXOR (bagua.torch_api.communication.ReduceOp attribute)
(bagua.torch_api.ReduceOp attribute)
ByteGradAlgorithm (class in bagua.torch_api.algorithms.bytegrad)
ByteGradAlgorithmImpl (class in bagua.torch_api.algorithms.bytegrad)
bytes() (bagua.torch_api.bucket.BaguaBucket method)
C
cache_loader (bagua.torch_api.contrib.cached_dataset.CachedDataset attribute)
(bagua.torch_api.contrib.CachedDataset attribute)
CachedDataset (class in bagua.torch_api.contrib)
(class in bagua.torch_api.contrib.cached_dataset)
CacheLoader (class in bagua.torch_api.contrib)
(class in bagua.torch_api.contrib.cache_loader)
check_flatten() (bagua.torch_api.bucket.BaguaBucket method)
clear() (bagua.torch_api.contrib.utils.store.Store method)
clear_ops() (bagua.torch_api.bucket.BaguaBucket method)
ClusterStore (class in bagua.torch_api.contrib.utils.store)
convert_sync_batchnorm() (bagua.torch_api.contrib.sync_batchnorm.SyncBatchNorm class method)
D
DecentralizedAlgorithm (class in bagua.torch_api.algorithms.decentralized)
DecentralizedAlgorithmImpl (class in bagua.torch_api.algorithms.decentralized)
E
ensure_bagua_tensor() (bagua.torch_api.BaguaTensor method)
(bagua.torch_api.tensor.BaguaTensor method)
exp_selection_uniform_map (in module bagua.torch_api.model_parallel.moe.sharded_moe)
Experts (class in bagua.torch_api.model_parallel.moe.experts)
F
flattened_tensor() (bagua.torch_api.bucket.BaguaBucket method)
forward() (bagua.torch_api.contrib.sync_batchnorm.SyncBatchNorm method)
(bagua.torch_api.model_parallel.moe.experts.Experts method)
(bagua.torch_api.model_parallel.moe.layer.MoE method)
(bagua.torch_api.model_parallel.moe.MoE method)
(bagua.torch_api.model_parallel.moe.sharded_moe.MOELayer method)
(bagua.torch_api.model_parallel.moe.sharded_moe.TopKGate method)
from_torch_group() (in module bagua.torch_api.communication)
fuse_optimizer() (in module bagua.torch_api.contrib)
(in module bagua.torch_api.contrib.fuse.optimizer)
fuse_step() (in module bagua.torch_api.contrib.fuse.optimizer)
G
gather() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
gather_inplace() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
get() (bagua.torch_api.contrib.cache_loader.CacheLoader method)
(bagua.torch_api.contrib.CacheLoader method)
(bagua.torch_api.contrib.utils.store.Store method)
get_backend() (in module bagua.torch_api)
get_default_bucket_size() (in module bagua.torch_api.env)
get_global_communicator() (bagua.torch_api.communication.BaguaProcessGroup method)
get_inter_node_communicator() (bagua.torch_api.communication.BaguaProcessGroup method)
get_intra_node_communicator() (bagua.torch_api.communication.BaguaProcessGroup method)
get_local_rank() (in module bagua.torch_api)
(in module bagua.torch_api.env)
get_local_size() (in module bagua.torch_api)
(in module bagua.torch_api.env)
get_rank() (in module bagua.torch_api)
(in module bagua.torch_api.env)
get_world_size() (in module bagua.torch_api)
(in module bagua.torch_api.env)
GradientAllReduceAlgorithm (class in bagua.torch_api.algorithms.gradient_allreduce)
GradientAllReduceAlgorithmImpl (class in bagua.torch_api.algorithms.gradient_allreduce)
gumbel_map (in module bagua.torch_api.model_parallel.moe.sharded_moe)
gumbel_rsample() (in module bagua.torch_api.model_parallel.moe.sharded_moe)
I
init_backward_hook() (bagua.torch_api.algorithms.AlgorithmImpl method)
(bagua.torch_api.algorithms.base.AlgorithmImpl method)
init_forward_pre_hook() (bagua.torch_api.algorithms.AlgorithmImpl method)
(bagua.torch_api.algorithms.base.AlgorithmImpl method)
init_operations() (bagua.torch_api.algorithms.AlgorithmImpl method)
(bagua.torch_api.algorithms.base.AlgorithmImpl method)
init_post_backward_hook() (bagua.torch_api.algorithms.AlgorithmImpl method)
(bagua.torch_api.algorithms.base.AlgorithmImpl method)
init_post_optimizer_step_hook() (bagua.torch_api.algorithms.AlgorithmImpl method)
(bagua.torch_api.algorithms.base.AlgorithmImpl method)
init_process_group() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
init_tensors() (bagua.torch_api.algorithms.AlgorithmImpl method)
(bagua.torch_api.algorithms.base.AlgorithmImpl method)
is_bagua_tensor() (bagua.torch_api.BaguaTensor method)
(bagua.torch_api.tensor.BaguaTensor method)
is_fused_optimizer() (in module bagua.torch_api.contrib.fuse.optimizer)
is_initialized() (in module bagua.torch_api.communication)
is_moe_param() (in module bagua.torch_api.model_parallel.moe)
(in module bagua.torch_api.model_parallel.moe.utils)
L
load_checkpoint() (in module bagua.torch_api.checkpoint)
(in module bagua.torch_api.checkpoint.checkpointing)
LoadBalancingDistributedBatchSampler (class in bagua.torch_api.contrib)
(class in bagua.torch_api.contrib.load_balancing_data_loader)
LoadBalancingDistributedSampler (class in bagua.torch_api.contrib)
(class in bagua.torch_api.contrib.load_balancing_data_loader)
LooseVersion (in module bagua.torch_api.tensor)
LowPrecisionDecentralizedAlgorithm (class in bagua.torch_api.algorithms.decentralized)
LowPrecisionDecentralizedAlgorithmImpl (class in bagua.torch_api.algorithms.decentralized)
M
MAX (bagua.torch_api.communication.ReduceOp attribute)
(bagua.torch_api.ReduceOp attribute)
mget() (bagua.torch_api.contrib.utils.store.Store method)
MIN (bagua.torch_api.communication.ReduceOp attribute)
(bagua.torch_api.ReduceOp attribute)
module
bagua
bagua.torch_api
bagua.torch_api.algorithms
bagua.torch_api.algorithms.async_model_average
bagua.torch_api.algorithms.base
bagua.torch_api.algorithms.bytegrad
bagua.torch_api.algorithms.decentralized
bagua.torch_api.algorithms.gradient_allreduce
bagua.torch_api.algorithms.q_adam
bagua.torch_api.bucket
bagua.torch_api.checkpoint
bagua.torch_api.checkpoint.checkpointing
bagua.torch_api.communication
bagua.torch_api.contrib
bagua.torch_api.contrib.cache_loader
bagua.torch_api.contrib.cached_dataset
bagua.torch_api.contrib.fuse
bagua.torch_api.contrib.fuse.optimizer
bagua.torch_api.contrib.load_balancing_data_loader
bagua.torch_api.contrib.sync_batchnorm
bagua.torch_api.contrib.utils
bagua.torch_api.contrib.utils.redis_store
bagua.torch_api.contrib.utils.store
bagua.torch_api.distributed
bagua.torch_api.env
bagua.torch_api.model_parallel
bagua.torch_api.model_parallel.moe
bagua.torch_api.model_parallel.moe.experts
bagua.torch_api.model_parallel.moe.layer
bagua.torch_api.model_parallel.moe.sharded_moe
bagua.torch_api.model_parallel.moe.utils
bagua.torch_api.tensor
MoE (class in bagua.torch_api.model_parallel.moe)
(class in bagua.torch_api.model_parallel.moe.layer)
MOELayer (class in bagua.torch_api.model_parallel.moe.sharded_moe)
mset() (bagua.torch_api.contrib.utils.store.Store method)
multiplicative_jitter() (in module bagua.torch_api.model_parallel.moe.sharded_moe)
N
name (bagua.torch_api.bucket.BaguaBucket attribute)
need_reset() (bagua.torch_api.algorithms.AlgorithmImpl method)
(bagua.torch_api.algorithms.base.AlgorithmImpl method)
new_group() (in module bagua.torch_api.communication)
num_keys() (bagua.torch_api.contrib.cache_loader.CacheLoader method)
(bagua.torch_api.contrib.CacheLoader method)
(bagua.torch_api.contrib.utils.store.Store method)
P
PRODUCT (bagua.torch_api.communication.ReduceOp attribute)
(bagua.torch_api.ReduceOp attribute)
Q
QAdamAlgorithm (class in bagua.torch_api.algorithms.q_adam)
QAdamAlgorithmImpl (class in bagua.torch_api.algorithms.q_adam)
QAdamOptimizer (class in bagua.torch_api.algorithms.q_adam)
R
recv() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
RedisStore (class in bagua.torch_api.contrib.utils.redis_store)
reduce() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
reduce_inplace() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
reduce_scatter() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
reduce_scatter_inplace() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
ReduceOp (class in bagua.torch_api)
(class in bagua.torch_api.communication)
reify() (bagua.torch_api.algorithms.Algorithm method)
(bagua.torch_api.algorithms.base.Algorithm method)
resume() (bagua.torch_api.algorithms.async_model_average.AsyncModelAverageAlgorithmImpl method)
S
save_checkpoint() (in module bagua.torch_api.checkpoint)
(in module bagua.torch_api.checkpoint.checkpointing)
scatter() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
scatter_inplace() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
send() (in module bagua.torch_api)
(in module bagua.torch_api.communication)
set() (bagua.torch_api.contrib.utils.store.Store method)
set_epoch() (bagua.torch_api.contrib.load_balancing_data_loader.LoadBalancingDistributedBatchSampler method)
(bagua.torch_api.contrib.load_balancing_data_loader.LoadBalancingDistributedSampler method)
(bagua.torch_api.contrib.LoadBalancingDistributedBatchSampler method)
(bagua.torch_api.contrib.LoadBalancingDistributedSampler method)
shutdown() (bagua.torch_api.contrib.utils.store.Store method)
status() (bagua.torch_api.contrib.utils.store.Store method)
step() (bagua.torch_api.algorithms.q_adam.QAdamOptimizer method)
Store (class in bagua.torch_api.contrib.utils.store)
SUM (bagua.torch_api.communication.ReduceOp attribute)
(bagua.torch_api.ReduceOp attribute)
SyncBatchNorm (class in bagua.torch_api.contrib.sync_batchnorm)
T
tensors (bagua.torch_api.bucket.BaguaBucket attribute)
tensors_to_buckets() (bagua.torch_api.algorithms.AlgorithmImpl method)
(bagua.torch_api.algorithms.base.AlgorithmImpl method)
to_bagua_tensor() (bagua.torch_api.BaguaTensor method)
(bagua.torch_api.tensor.BaguaTensor method)
top1gating() (in module bagua.torch_api.model_parallel.moe.sharded_moe)
top2gating() (in module bagua.torch_api.model_parallel.moe.sharded_moe)
TopKGate (class in bagua.torch_api.model_parallel.moe.sharded_moe)
U
uniform_map (in module bagua.torch_api.model_parallel.moe.sharded_moe)
unused (in module bagua.torch_api.contrib.sync_batchnorm)
V
version (in module bagua.torch_api)
W
wg (bagua.torch_api.model_parallel.moe.sharded_moe.TopKGate attribute)
with_bagua() (bagua.torch_api.BaguaModule method)
(bagua.torch_api.distributed.BaguaModule method)
Read the Docs
v: v0.8.2
Versions
latest
stable
v0.8.2
v0.8.1.post1
v0.8.1
v0.8.0
v0.7.0
v0.6.3
v0.6.2
v0.6.1
v0.6.0
0.5.0
0.4.0
Downloads
On Read the Docs
Project Home
Builds
Free document hosting provided by
Read the Docs
.