Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
机器未来
Paddle
提交
0b8067c0
P
Paddle
项目概览
机器未来
/
Paddle
与 Fork 源项目一致
Fork自
PaddlePaddle / Paddle
通知
1
Star
1
Fork
0
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
1
列表
看板
标记
里程碑
合并请求
0
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
P
Paddle
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
1
Issue
1
列表
看板
标记
里程碑
合并请求
0
合并请求
0
Pages
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
提交
Issue看板
未验证
提交
0b8067c0
编写于
6年前
作者:
W
Wu Yi
提交者:
GitHub
6年前
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
fix dist train reduce mode (#13068)
* fix dist train reduce mode * fix previous fix
上级
823c4f87
develop
2.0.1-rocm-post
Ligoml-patch-1
OliverLPH-patch-1
OliverLPH-patch-2
PaddlePM-patch-1
PaddlePM-patch-2
ZHUI-patch-1
add_default_att
add_model_benchmark_ci
add_some_yaml_config
addfile
all_new_design_exec
ascendrc
ascendrelease
cherry_undefined_var
compile_windows
delete_2.0.1-rocm-post
delete_add_default_att
delete_all_new_design_exec
delete_ascendrc
delete_compile_windows
delete_delete_addfile
delete_disable_iterable_dataset_unittest
delete_fix_dataloader_memory_leak
delete_fix_imperative_dygraph_error
delete_fix_retry_ci
delete_fix_undefined_var
delete_improve_sccache
delete_incubate/lite
delete_paddle_tiny_install
delete_paralleltest
delete_prv-disable-more-cache
delete_revert-31068-fix_conv3d_windows
delete_revert-31562-mean
delete_revert-33630-bug-fix
delete_revert-34159-add_npu_bce_logical_dev
delete_revert-34910-spinlocks_for_allocator
delete_revert-35069-revert-34910-spinlocks_for_allocator
delete_revert-36057-dev/read_flags_in_ut
dingjiaweiww-patch-1
disable_iterable_dataset_unittest
dy2static
enable_eager_model_test
final_state_gen_python_c
final_state_intermediate
fix-numpy-issue
fix_concat_slice
fix_dataloader_memory_leak
fix_imperative_dygraph_error
fix_npu_ci
fix_op_flops
fix_retry_ci
fix_rnn_docs
fix_tensor_type
fix_undefined_var
fixiscan
fixiscan1
fixiscan2
fixiscan3
github/fork/123malin/netifaces
github/fork/123malin/tdm_abacus
github/fork/AshburnLee/dev_unique
github/fork/ForFishes/fix_memory_matmul
github/fork/ForFishes/rm_fluid
github/fork/LielinJiang/move-2.0-api
github/fork/LielinJiang/visual-dl-cb
github/fork/LiuChiachi/add-transformer-generate-square-subsequent-mask-api
github/fork/LiuChiachi/fix-example-code-for-hapi-Model
github/fork/LiuChiachi/remove-input-requirment-in-dygraph-Model
github/fork/MrChengmo/fix_ps_profiler
github/fork/MrChengmo/update_ps_heter
github/fork/PWhiddy/patch-1
github/fork/Shixiaowei02/dev/save_load_upgrade
github/fork/TCChenlong/fix_hapi
github/fork/TCChenlong/fix_inden
github/fork/Thunderbrook/xpu_slice
github/fork/XieYunshen/disable_ut_test_parallel_executor_fetch_isolated_var
github/fork/XieYunshen/disable_ut_test_parallel_executor_fetch_isolated_var_2
github/fork/XieYunshen/disable_ut_test_parallel_executor_fetch_isolated_var_3
github/fork/XieYunshen/timeout_20S_ut
github/fork/ZeyuChen/remove-nltk
github/fork/arlesniak/arlesniak/selective__mkldnn_flags
github/fork/baiyfbupt/code_doc_mig
github/fork/chalsliu/set_timeout
github/fork/chen-zhiyu/develop
github/fork/chenwhql/ci/try_to_find_test_buffer_shared_memory_reuse_pass_error
github/fork/chenwhql/dygraph/remove_scale_loss_and_apply_collective_grads
github/fork/chenwhql/saveload/add_get_inference_program
github/fork/chenwhql/saveload/remove_save_load_config
github/fork/cryoco/pass-compatibility-trt
github/fork/danleifeng/isempty_api2.0
github/fork/frankwhzhang/api_transfer
github/fork/hbwx24/error_msg/cuda_kernel_error_msg
github/fork/heavengate/cherry_yolo_box
github/fork/heavengate/update_yolo_box
github/fork/iclementine/rnn_fix
github/fork/iducn/testestse
github/fork/jczaja/prv-25537-fix
github/fork/jeff41404/release/1.8
github/fork/jiweibo/api_2.0
github/fork/jiweibo/fix_lite_resnet50_test
github/fork/juncaipeng/fix_doc_1
github/fork/lfchener/sample_code
github/fork/littletomatodonkey/fix_reg_doc
github/fork/liym27/dy2stat_update_assign_to_rc20
github/fork/luotao1/profiler_ut
github/fork/mapingshuo/add_wait
github/fork/mapingshuo/doc_2.0
github/fork/mapingshuo/zero-0.5
github/fork/miraiwk/dev
github/fork/pangyoki/add-Categorical-class-branch
github/fork/pangyoki/add-multinomial-op-branch
github/fork/pangyoki/fix-test_distritbution-CI
github/fork/qjing666/doublegrad
github/fork/qjing666/fix_hdfs_download
github/fork/sandyhouse/add_gather_etc
github/fork/sandyhouse/add_send_recv_alltoall_etc
github/fork/sandyhouse/pipeline_exe_run
github/fork/seiriosPlus/feature/large_scale_kv_save_delta
github/fork/seiriosPlus/fix/paddle_errors_fix
github/fork/seiriosPlus/fix/paddle_op_errors
github/fork/shangzhizhou/fix_test_activation_op_random_bug
github/fork/smallv0221/yxp0924
github/fork/smallv0221/yxp0925
github/fork/swtkiwi/del-matplotlib
github/fork/tianshuo78520a/kunlun_test
github/fork/tianshuo78520a/update_dockerfile
github/fork/wanghaoshuang/bert_fuse
github/fork/wanghaoshuang/label_smooth
github/fork/wanghuancoder/develop_CUDASynchronize
github/fork/wanghuancoder/develop_Layer_doc
github/fork/wanghuancoder/develop_ParameterList_doc
github/fork/wanghuancoder/develop_Sequential_doc
github/fork/wanghuancoder/develop_bilinear_tensor_product
github/fork/wanghuancoder/develop_coverage_build_sh
github/fork/wanghuancoder/develop_in_dynamic_mode_doc
github/fork/wanghuancoder/develop_unique_name_doc
github/fork/wangxicoding/fleet_meta_combine
github/fork/wawltor/error_message_fix_5
github/fork/willthefrog/remove_l2_norm
github/fork/windstamp/momentum_op
github/fork/windstamp/mv_op_5
github/fork/windstamp/normal_api
github/fork/wojtuss/wojtuss/fusion_gru_quantization
github/fork/wojtuss/wojtuss/quantization-with-shift
github/fork/wzzju/fix_err_info
github/fork/wzzju/pure_fp16
github/fork/xiemoyuan/op_error_message
github/fork/xiemoyuan/optimize_error_message
github/fork/yaoxuefeng6/fix_doc
github/fork/yaoxuefeng6/mod_dataset_v2
github/fork/yongqiangma/lod
github/fork/ysh329/fix-clip-by-norm-error
github/fork/ysh329/fix-error-clip-by-value
github/fork/yukavio/error_info
github/fork/zhangting2020/conv_filter_grad
github/fork/zhangting2020/is_compile_with_cuda
github/fork/zhangting2020/place_doc
github/fork/zhangting2020/program
github/fork/zhhsplendid/fix_any
github/fork/zhhsplendid/refine_api2
github/fork/zhhsplendid/refine_api2_test
github/fork/zhhsplendid/refine_api_test_ptb_lm
github/fork/zhhsplendid/refine_api_test_resnet
github/fork/zhhsplendid/refine_api_test_simnet
github/fork/zhiqiu/dev/refine_initializer
github/fork/zhiqiu/dev/remove_inplace_argument
github/fork/zlsh80826/nvinfer_plugin_var_len_cuda11
improve_sccache
incubate/infrt
incubate/lite
inplace_addto
make_flag_adding_easier
move_embedding_to_phi
move_histogram_to_pten
move_sgd_to_phi
move_slice_to_pten
move_temporal_shift_to_phi
move_yolo_box_to_phi
npu_fix_alloc
numel
paddle_tiny_install
paralleltest
preln_ernie
prv-disable-more-cache
prv-md-even-more
prv-onednn-2.5
pten_tensor_refactor
release/1.0.0
release/1.1
release/1.2
release/1.3
release/1.4
release/1.5
release/1.6
release/1.7
release/1.8
release/2.0
release/2.0-alpha
release/2.0-beta
release/2.0-rc
release/2.0-rc1
release/2.1
release/2.2
release/2.3
release/2.3-fc-ernie-fix
release/2.4
release/lite-0.1
revert-24981-add_device_attr_for_regulization
revert-26856-strategy_example2
revert-27520-disable_pr
revert-31068-fix_conv3d_windows
revert-31562-mean
revert-32290-develop-hardlabel
revert-33037-forci
revert-33475-fix_cifar_label_dimension
revert-33630-bug-fix
revert-34159-add_npu_bce_logical_dev
revert-34406-add_copy_from_tensor
revert-34910-spinlocks_for_allocator
revert-35069-revert-34910-spinlocks_for_allocator
revert-36057-dev/read_flags_in_ut
revert-36201-refine_fast_threaded_ssa_graph_executor
revert-36985-add_license
revert-37318-refactor_dygraph_to_eager
revert-37926-eager_coreops_500
revert-37956-revert-37727-pylayer_support_tuple
revert-38100-mingdong
revert-38301-allocation_rearrange_pr
revert-38703-numpy_bf16_package_reupload
revert-38732-remove_useless_header_in_elementwise_mul_grad
revert-38959-Reduce_Grad
revert-39143-adjust_empty
revert-39227-move_trace_op_to_pten
revert-39268-dev/remove_concat_fluid_kernel
revert-40170-support_partial_grad
revert-41056-revert-40727-move_some_activaion_to_phi
revert-41065-revert-40993-mv_ele_floordiv_pow
revert-41068-revert-40790-phi_new
revert-41944-smaller_inference_api_test
revert-42149-do-not-reset-default-stream-for-stream-safe-cuda-allocator
revert-43155-fix_ut_tempfile
revert-43882-revert-41944-smaller_inference_api_test
revert-45808-phi/simplify_size_op
revert-46827-deform_comment
rocm_dev_0217
support_weight_transpose
test_benchmark_ci
test_feature_precision_test_c
test_model_benchmark
test_model_benchmark_ci
zhiqiu-patch-1
v2.4.0-rc0
v2.3.2
v2.3.1
v2.3.0
v2.3.0-rc0
v2.2.2
v2.2.1
v2.2.0
v2.2.0-rc0
v2.2.0-bak0
v2.1.3
v2.1.2
v2.1.1
v2.1.0
v2.1.0-rc0
v2.0.2
v2.0.1
v2.0.0
v2.0.0-rc1
v2.0.0-rc0
v2.0.0-beta0
v2.0.0-alpha0
v1.8.5
v1.8.4
v1.8.3
v1.8.2
v1.8.1
v1.8.0
v1.7.2
v1.7.1
v1.7.0
v1.6.3
v1.6.2
v1.6.1
v1.6.0
v1.6.0-rc0
v1.5.2
v1.5.1
v1.5.0
v1.4.1
v1.4.0
v1.3.2
v1.3.1
v1.3.0
v1.2.1
v1.2.0
v1.1.0
v1.0.2
v1.0.1
v1.0.0
v1.0.0-rc0
lite-v0.1
无相关合并请求
变更
4
隐藏空白更改
内联
并排
Showing
4 changed file
with
79 addition
and
21 deletion
+79
-21
paddle/fluid/framework/details/multi_devices_graph_pass.cc
paddle/fluid/framework/details/multi_devices_graph_pass.cc
+1
-1
python/paddle/fluid/tests/unittests/test_dist_base.py
python/paddle/fluid/tests/unittests/test_dist_base.py
+38
-14
python/paddle/fluid/tests/unittests/test_dist_mnist.py
python/paddle/fluid/tests/unittests/test_dist_mnist.py
+21
-0
python/paddle/fluid/transpiler/distribute_transpiler.py
python/paddle/fluid/transpiler/distribute_transpiler.py
+19
-6
未找到文件。
paddle/fluid/framework/details/multi_devices_graph_pass.cc
浏览文件 @
0b8067c0
...
...
@@ -744,7 +744,7 @@ void MultiDevSSAGraphBuilder::CreateDistTrainOp(ir::Graph *result,
.
emplace
(
varname
,
op_dev_id
);
}
}
else
{
PADDLE_
ENFORCE
(
PADDLE_
THROW
(
"the distribute training related op should be in [split_byref, "
"concat]."
);
}
...
...
This diff is collapsed.
Click to expand it.
python/paddle/fluid/tests/unittests/test_dist_base.py
浏览文件 @
0b8067c0
...
...
@@ -82,8 +82,18 @@ class TestDistRunnerBase(object):
strategy
=
fluid
.
ExecutionStrategy
()
strategy
.
num_threads
=
1
strategy
.
allow_op_delay
=
False
build_stra
=
fluid
.
BuildStrategy
()
if
args
.
use_reduce
:
build_stra
.
reduce_strategy
=
fluid
.
BuildStrategy
.
ReduceStrategy
.
Reduce
else
:
build_stra
.
reduce_strategy
=
fluid
.
BuildStrategy
.
ReduceStrategy
.
AllReduce
exe
=
fluid
.
ParallelExecutor
(
True
,
loss_name
=
avg_cost
.
name
,
exec_strategy
=
strategy
)
True
,
loss_name
=
avg_cost
.
name
,
exec_strategy
=
strategy
,
build_strategy
=
build_stra
)
feed_var_list
=
[
var
for
var
in
trainer_prog
.
global_block
().
vars
.
values
()
...
...
@@ -123,6 +133,7 @@ def runtime_main(test_class):
'--current_endpoint'
,
type
=
str
,
required
=
False
,
default
=
""
)
parser
.
add_argument
(
'--sync_mode'
,
action
=
'store_true'
)
parser
.
add_argument
(
'--mem_opt'
,
action
=
'store_true'
)
parser
.
add_argument
(
'--use_reduce'
,
action
=
'store_true'
)
args
=
parser
.
parse_args
()
...
...
@@ -149,20 +160,25 @@ class TestDistBase(unittest.TestCase):
self
.
_python_interp
=
"python"
self
.
_sync_mode
=
True
self
.
_mem_opt
=
False
self
.
_use_reduce
=
False
self
.
_setup_config
()
def
start_pserver
(
self
,
model_file
,
check_error_log
):
ps0_ep
,
ps1_ep
=
self
.
_ps_endpoints
.
split
(
","
)
ps_cmd
=
"%s %s --role pserver --endpoints %s --trainer_id 0 --current_endpoint %s --trainers %d --is_dist %s %s"
sync_mode_str
=
"--sync_mode"
if
self
.
_sync_mode
else
""
mem_opt_str
=
"--mem_opt"
if
self
.
_mem_opt
else
""
ps_cmd
=
"%s %s --role pserver --endpoints %s --trainer_id 0 --current_endpoint %s --trainers %d --is_dist"
ps0_cmd
=
ps_cmd
%
\
(
self
.
_python_interp
,
model_file
,
self
.
_ps_endpoints
,
ps0_ep
,
self
.
_trainers
,
sync_mode_str
,
mem_opt_str
)
self
.
_trainers
)
ps1_cmd
=
ps_cmd
%
\
(
self
.
_python_interp
,
model_file
,
self
.
_ps_endpoints
,
ps1_ep
,
self
.
_trainers
,
sync_mode_str
,
mem_opt_str
)
self
.
_trainers
)
if
self
.
_sync_mode
:
ps0_cmd
+=
" --sync_mode"
ps1_cmd
+=
" --sync_mode"
if
self
.
_mem_opt
:
ps0_cmd
+=
" --mem_opt"
ps1_cmd
+=
" --mem_opt"
ps0_pipe
=
subprocess
.
PIPE
ps1_pipe
=
subprocess
.
PIPE
...
...
@@ -242,17 +258,23 @@ class TestDistBase(unittest.TestCase):
self
.
_wait_ps_ready
(
ps1
.
pid
)
ps0_ep
,
ps1_ep
=
self
.
_ps_endpoints
.
split
(
","
)
tr_cmd
=
"%s %s --role trainer --endpoints %s --trainer_id %d --current_endpoint %s --trainers %d --is_dist %s %s"
sync_mode_str
=
"--sync_mode"
if
self
.
_sync_mode
else
""
mem_opt_str
=
"--mem_opt"
if
self
.
_mem_opt
else
""
tr_cmd
=
"%s %s --role trainer --endpoints %s --trainer_id %d --current_endpoint %s --trainers %d --is_dist"
tr0_cmd
=
tr_cmd
%
\
(
self
.
_python_interp
,
model_file
,
self
.
_ps_endpoints
,
0
,
ps0_ep
,
self
.
_trainers
,
sync_mode_str
,
mem_opt_str
)
0
,
ps0_ep
,
self
.
_trainers
)
tr1_cmd
=
tr_cmd
%
\
(
self
.
_python_interp
,
model_file
,
self
.
_ps_endpoints
,
1
,
ps1_ep
,
self
.
_trainers
,
sync_mode_str
,
mem_opt_str
)
1
,
ps1_ep
,
self
.
_trainers
)
if
self
.
_sync_mode
:
tr0_cmd
+=
" --sync_mode"
tr1_cmd
+=
" --sync_mode"
if
self
.
_mem_opt
:
tr0_cmd
+=
" --mem_opt"
tr1_cmd
+=
" --mem_opt"
if
self
.
_use_reduce
:
tr0_cmd
+=
" --use_reduce"
tr1_cmd
+=
" --use_reduce"
env0
=
{
"CUDA_VISIBLE_DEVICES"
:
"0"
}
env1
=
{
"CUDA_VISIBLE_DEVICES"
:
"1"
}
...
...
@@ -303,6 +325,8 @@ class TestDistBase(unittest.TestCase):
# FIXME: use terminate() instead of sigkill.
os
.
kill
(
ps0
.
pid
,
signal
.
SIGKILL
)
os
.
kill
(
ps1
.
pid
,
signal
.
SIGKILL
)
ps0
.
terminate
()
ps1
.
terminate
()
ps0
.
wait
()
ps1
.
wait
()
FNULL
.
close
()
...
...
This diff is collapsed.
Click to expand it.
python/paddle/fluid/tests/unittests/test_dist_mnist.py
浏览文件 @
0b8067c0
...
...
@@ -20,6 +20,7 @@ from test_dist_base import TestDistBase
class
TestDistMnist2x2
(
TestDistBase
):
def
_setup_config
(
self
):
self
.
_sync_mode
=
True
self
.
_use_reduce
=
False
def
test_se_resnext
(
self
):
self
.
check_with_place
(
"dist_mnist.py"
,
delta
=
1e-7
)
...
...
@@ -37,10 +38,30 @@ class TestDistMnist2x2WithMemopt(TestDistBase):
class
TestDistMnistAsync
(
TestDistBase
):
def
_setup_config
(
self
):
self
.
_sync_mode
=
False
self
.
_use_reduce
=
False
def
test_se_resnext
(
self
):
self
.
check_with_place
(
"dist_mnist.py"
,
delta
=
200
)
# FIXME(typhoonzero): enable these tests once we have 4
# 4 GPUs on CI machine, and the base class should be updated.
#
# class TestDistMnist2x2ReduceMode(TestDistBase):
# def _setup_config(self):
# self._sync_mode = True
# self._use_reduce = True
# def test_se_resnext(self):
# self.check_with_place("dist_mnist.py", delta=1e-7)
# class TestDistMnistAsyncReduceMode(TestDistBase):
# def _setup_config(self):
# self._sync_mode = False
# self._use_reduce = True
# def test_se_resnext(self):
# self.check_with_place("dist_mnist.py", delta=200)
if
__name__
==
"__main__"
:
unittest
.
main
()
This diff is collapsed.
Click to expand it.
python/paddle/fluid/transpiler/distribute_transpiler.py
浏览文件 @
0b8067c0
...
...
@@ -273,6 +273,10 @@ class DistributeTranspiler(object):
name
=
framework
.
generate_control_dev_var_name
())
grad_name_to_send_dummy_out
[
grad_varname
]
=
dummy_output
# get send op_role_var, if not splited, the grad should have .trainer suffix
# if splited, grad should be the original grad var name (split_by_ref and send
# will be on the same place). ParallelExecutor
# will use op_role_var to get expected device place to run this op.
program
.
global_block
().
_insert_op
(
index
=
index
+
1
,
type
=
"send"
,
...
...
@@ -281,8 +285,10 @@ class DistributeTranspiler(object):
attrs
=
{
"epmap"
:
eplist
,
RPC_OP_ROLE_ATTR_NAME
:
RPC_OP_ROLE_ATTR_VALUE
,
OP_ROLE_VAR_ATTR_NAME
:
[
self
.
grad_name_to_param_name
[
grad_varname
],
grad_varname
],
OP_ROLE_VAR_ATTR_NAME
:
[
self
.
grad_name_to_param_name
[
grad_varname
],
splited_grad_varname
],
"sync_mode"
:
not
self
.
sync_mode
,
})
for
_
,
var
in
enumerate
(
splited_vars
):
...
...
@@ -326,6 +332,15 @@ class DistributeTranspiler(object):
recv_dep_in
=
grad_name_to_send_dummy_out
[
self
.
param_name_to_grad_name
[
param_varname
]]
all_recv_outputs
.
extend
(
splited_var
)
# get recv op_role_var, if not splited, the grad should have .trainer suffix
# if splited, grad should be the original grad var name. ParallelExecutor
# will use op_role_var to get expected device place to run this op.
orig_grad_name
=
self
.
param_name_to_grad_name
[
param_varname
]
recv_op_role_var_name
=
orig_grad_name
splited_trainer_grad
=
self
.
grad_var_mapping
[
orig_grad_name
]
if
len
(
splited_trainer_grad
)
==
1
:
recv_op_role_var_name
=
splited_trainer_grad
[
0
].
name
program
.
global_block
().
append_op
(
type
=
"recv"
,
inputs
=
{
"X"
:
[
recv_dep_in
]},
...
...
@@ -333,10 +348,8 @@ class DistributeTranspiler(object):
attrs
=
{
"epmap"
:
eps
,
RPC_OP_ROLE_ATTR_NAME
:
RPC_OP_ROLE_ATTR_VALUE
,
OP_ROLE_VAR_ATTR_NAME
:
[
param_varname
,
self
.
param_name_to_grad_name
[
param_varname
]
],
OP_ROLE_VAR_ATTR_NAME
:
[
param_varname
,
recv_op_role_var_name
],
"sync_mode"
:
not
self
.
sync_mode
})
...
...
This diff is collapsed.
Click to expand it.
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录
反馈
建议
客服
返回
顶部