PaddlePaddle（飞桨） v1.6 深度学习平台使用文档
首页
白天
夜间
下载
阅读记录
书签管理
我的书签
添加书签
移除书签
编辑文档
memory_optimize
来源 1
浏览
456
扫码
打印
2020-05-04 12:12:47
memory_optimize
上一篇:
下一篇:
发布点评
进阶使用
模型压缩工具库
优秀实践
分布式CPU训练优秀实践
单机训练优秀实践
存储分配与优化
分布式GPU训练优秀实践
如何贡献代码
提交PR注意事项
本地开发指南
FAQ
新增OP
如何在框架外部自定义C++ OP
如何写新的C++ OP
如何写新的Python OP
C++ OP相关注意事项
性能调优
timeline工具简介
CPU性能调优
堆内存分析和优化
预测部署
移动端部署
Paddle-Lite
服务器端部署
安装与编译C++预测库
使用Paddle-TensorRT库预测
Python 预测 API介绍
C++ 预测 API介绍
C 预测 API介绍
安装与编译Windows预测库
如何贡献文档
设计思想
新手入门
深度学习基础教程
线性回归
机器翻译
数字识别
词向量
语义角色标注
个性化推荐
生成对抗网络
情感分析
图像分类
编程指南
快速开始
安装说明
Ubuntu下安装
附录
从源码编译
Ubuntu下从源码编译
CentOS下从源码编译
MacOS下从源码编译
Windows下从源码编译
CentOS下安装
MacOS下安装
使用Docker安装
使用conda安装
Windows下安装
使用指南
准备数据
同步数据读取
数据预处理工具
异步数据读取
准备步骤
模型库
动态图机制-DyGraph
训练神经网络
多机训练
分布式训练使用手册
1. 总体概览
ELASTIC CTR
3. 分布式训练+Serving方案一键部署
使用FleetAPI进行分布式训练
5. 二次开发指南
2. 前置需求
分布式训练快速开始
4. 查看结果
注释
单机训练
训练过程中评测模型
模型/变量的保存、载入与增量训练
模型评估/调试
VisualDL 工具
VisualDL 工具简介
VisualDL 使用指南
模型评估
配置简单的网络
基本概念
Tensor
Executor
Program
Variable
LoDTensor
Operator
FAQ
预测引擎
网络搭建及训练
安装与编译
环境变量FLAGS
其他
存储管理
cudnn
执行器
调试
分布式
数值计算
设备管理
API Reference
fluid.dygraph
PolynomialDecay
save_dygraph
BilinearTensorProduct
InverseTimeDecay
PRelu
GRUUnit
to_variable
CosineDecay
ExponentialDecay
load_dygraph
Embedding
GroupNorm
SpectralNorm
NaturalExpDecay
Pool2D
NoamDecay
TreeConv
Conv3D
no_grad
NCE
Conv2D
Conv2DTranspose
guard
PiecewiseDecay
BatchNorm
FC
BackwardStrategy
Layer
LayerNorm
Conv3DTranspose
API功能分类
数据并行执行引擎
模型参数
模型保存与加载
Caffe-Fluid常用层对应表
神经网络层
池化
序列
数据输入输出
学习率调度器
控制流
图像检测
张量
使用DataFeeder传入训练/预测数据
数学操作
损失函数
稀疏更新
激活函数
卷积
优化器
TensorFlow-Fluid常用接口对应表
预测引擎
反向传播
执行引擎
基础概念
评价指标
CompiledProgram
分布式训练
分布式训练reader准备
分布式同步训练
大规模稀疏特征模型训练
分布式异步训练
复杂网络
fluid.dataset
InMemoryDataset
DatasetFactory
QueueDataset
fluid.optimizer
AdadeltaOptimizer
RecomputeOptimizer
ModelAverage
Adam
Adadelta
Adamax
AdagradOptimizer
DecayedAdagradOptimizer
SGDOptimizer
ExponentialMovingAverage
AdamaxOptimizer
DecayedAdagrad
LarsMomentum
Adagrad
Momentum
LarsMomentumOptimizer
LambOptimizer
FtrlOptimizer
MomentumOptimizer
DGCMomentumOptimizer
RMSPropOptimizer
SGD
LookaheadOptimizer
AdamOptimizer
PipelineOptimizer
Ftrl
fluid.transpiler
release_memory
memory_optimize
DistributeTranspilerConfig
DistributeTranspiler
RoundRobin
HashName
dataset
imdb
imikolov
mnist
wmt14
wmt16
uci_housing
movielens
cifar
Conll05
sentiment
fluid.initializer
MSRAInitializer
Normal
TruncatedNormalInitializer
XavierInitializer
UniformInitializer
NormalInitializer
NumpyArrayInitializer
BilinearInitializer
TruncatedNormal
force_init_on_cpu
Bilinear
ConstantInitializer
Constant
init_on_cpu
Uniform
Xavier
MSRA
fluid.metrics
CompositeMetric
Auc
MetricBase
Recall
EditDistance
Precision
ChunkEvaluator
Accuracy
DetectionMAP
fluid.profiler
start_profiler
stop_profiler
profiler
cuda_profiler
reset_profiler
fluid.regularizer
L1Decay
L1DecayRegularizer
L2Decay
L2DecayRegularizer
fluid.unique_name
guard
generate
switch
fluid.io
cache
chain
compose
save_params
shuffle
load_params
load_vars
load
batch
firstn
save_vars
multiprocess_reader
map_readers
load_inference_model
DataLoader
save_persistables
xmap_readers
save
buffered
load_persistables
PyReader
save_inference_model
fluid.nets
sequence_conv_pool
simple_img_conv_pool
scaled_dot_product_attention
img_conv_group
glu
fluid.executor
global_scope
Executor
scope_guard
fluid
DistributeTranspiler
gradients
ParamAttr
create_random_int_lodtensor
name_scope
Executor
BuildStrategy
CUDAPlace
memory_optimize
CompiledProgram
create_lod_tensor
Variable
program_guard
one_hot
scope_guard
Program
DataFeeder
embedding
ParallelExecutor
is_compiled_with_cuda
data
cuda_places
ExecutionStrategy
Tensor
DataFeedDesc
CUDAPinnedPlace
LoDTensorArray
global_scope
LoDTensor
CPUPlace
DistributeTranspilerConfig
default_main_program
cuda_pinned_places
in_dygraph_mode
release_memory
cpu_places
WeightNormParamAttr
default_startup_program
fluid.layers
softmax_with_cross_entropy
squeeze
pool2d
assign
lod_reset
multiplex
array_write
piecewise_decay
box_clip
beam_search
expand
inverse_time_decay
maxout
im2sequence
logical_or
equal
round
reduce_any
rank
stanh
is_empty
double_buffer
bilinear_tensor_product
tensor_array_to_tensor
softplus
reduce_sum
gather_nd
edit_distance
conv2d_transpose
label_smooth
where
temporal_shift
Print
softshrink
conv3d_transpose
stack
ones_like
elementwise_min
pad_constant_like
logsigmoid
dropout
roi_align
eye
iou_similarity
warpctc
dynamic_decode
crop_tensor
RNNCell
noam_decay
sequence_expand_as
prior_box
floor
ones
spectral_norm
huber_loss
multiclass_nms
sequence_mask
data_norm
expand_as
reduce_prod
sequence_reverse
exponential_decay
py_reader
shape
create_tensor
layer_norm
MultivariateNormalDiag
affine_grid
asin
Decoder
cosine_decay
elementwise_add
has_inf
reorder_lod_tensor_by_rank
scale
sequence_unpad
gaussian_random_batch_size_like
merge_selected_rows
LSTMCell
chunk_eval
bpr_loss
unique
random_crop
generate_proposal_labels
elementwise_sub
retinanet_detection_output
lrn
less_equal
sampled_softmax_with_cross_entropy
While
shard_index
log
matmul
not_equal
sigmoid_cross_entropy_with_logits
roi_perspective_transform
diag
rsqrt
clip
grid_sampler
hard_swish
space_to_depth
tanh
leaky_relu
sequence_slice
sigmoid_focal_loss
create_py_reader_by_data
pixel_shuffle
distribute_fpn_proposals
cumsum
cos
less_than
square_error_cost
target_assign
read_file
brelu
pow
elementwise_mul
GRUCell
batch_norm
array_read
sequence_pool
hard_sigmoid
array_length
reduce_max
unique_with_counts
hard_shrink
sequence_reshape
create_parameter
rpn_target_assign
sums
box_coder
reduce_min
acos
elu
reverse
sin
selu
prroi_pool
concat
polygon_box_transform
elementwise_floordiv
pad2d
scatter_nd
reduce_all
zeros_like
filter_by_instag
continuous_value_model
rnn
crf_decoding
reciprocal
argsort
elementwise_max
sequence_concat
linspace
sqrt
deformable_conv
py_func
natural_exp_decay
fill_constant_batch_size_like
dice_loss
exp
sum
autoincreased_step_counter
split
nce
reduce_mean
lod_append
zeros
strided_slice
linear_chain_crf
topk
unfold
center_loss
elementwise_mod
hsigmoid
thresholded_relu
npair_loss
rank_loss
roi_pool
Categorical
log_loss
one_hot
data
density_prior_box
uniform_random
swish
group_norm
greater_equal
atan
scatter_nd_add
mean_iou
adaptive_pool2d
similarity_focus
detection_output
sigmoid
dynamic_gru
resize_bilinear
logical_not
argmax
reshape
conv3d
teacher_student_sigmoid_loss
sequence_first_step
gaussian_random
anchor_generator
collect_fpn_proposals
generate_mask_labels
cast
linear_lr_warmup
softsign
shuffle
accuracy
image_resize
instance_norm
gru_unit
increment
isfinite
create_global_var
beam_search_decode
Uniform
lstm_unit
mse_loss
ssd_loss
resize_nearest
sequence_scatter
mul
greater_than
yolov3_loss
embedding
sampling_id
yolo_box
abs
image_resize_short
generate_proposals
resize_trilinear
softmax
elementwise_div
sequence_pad
shuffle_channel
create_array
deformable_roi_pooling
retinanet_target_assign
has_nan
cos_sim
ctc_greedy_decoder
relu6
load
dynamic_lstm
square
conv2d
ceil
prelu
transpose
unsqueeze
scatter
add_position_encoding
auc
elementwise_pow
BeamSearchDecoder
hash
get_tensor_from_selected_rows
lstm
IfElse
uniform_random_batch_size_like
logical_and
tanh_shrink
DynamicRNN
gather_tree
sequence_enumerate
sequence_expand
sequence_softmax
mean
bipartite_match
polynomial_decay
dynamic_lstmp
box_decoder_and_assign
affine_channel
clip_by_norm
soft_relu
adaptive_pool3d
psroi_pool
smooth_l1
cross_entropy
l2_normalize
slice
fsp_matrix
unstack
pool3d
sign
relu
kldiv_loss
StaticRNN
argmin
range
multi_box_head
logical_xor
Switch
sequence_last_step
row_conv
fill_constant
margin_rank_loss
crop
Normal
pad
fc
sequence_conv
gather
flatten
fluid.clip
set_gradient_clip
GradientClipByGlobalNorm
GradientClipByNorm
GradientClipByValue
ErrorClipByValue
fluid.backward
append_backward
gradients
暂无相关搜索结果！
本文档使用
全库网
构建
×
思维导图备注
×
文章二维码
手机扫一扫，轻松掌上读
×
文档下载
请下载您需要的格式的文档，随时随地，享受汲取知识的乐趣！
PDF
文档
EPUB
文档
MOBI
文档
×
书签列表
×
阅读记录
阅读进度:
0.00%
(
0/0
)
重置阅读进度
我的书签
添加书签
移除书签