提交 070f1280 编写于 作者: D dongshuilong

fix distributied training bug for rec slim model

上级 0725a666
......@@ -87,9 +87,9 @@ DataLoader:
mean: [0., 0., 0.]
sampler:
name: DistributedRandomIdentitySampler
name: PKSampler
batch_size: 128
num_instances: 2
sample_per_id: 2
drop_last: False
shuffle: True
loader:
......
......@@ -92,9 +92,9 @@ DataLoader:
mean: [0., 0., 0.]
sampler:
name: DistributedRandomIdentitySampler
name: PKSampler
batch_size: 128
num_instances: 2
sample_per_id: 2
drop_last: False
shuffle: True
loader:
......
......@@ -91,9 +91,9 @@ DataLoader:
mean: [0., 0., 0.]
sampler:
name: DistributedRandomIdentitySampler
batch_size: 128
num_instances: 2
name: PKSampler
batch_size: 64
sample_per_id: 2
drop_last: False
shuffle: True
loader:
......
......@@ -95,9 +95,9 @@ DataLoader:
mean: [0., 0., 0.]
sampler:
name: DistributedRandomIdentitySampler
name: PKSampler
batch_size: 128
num_instances: 2
sample_per_id: 2
drop_last: False
shuffle: True
loader:
......
......@@ -94,9 +94,9 @@ DataLoader:
mean: [0., 0., 0.]
sampler:
name: DistributedRandomIdentitySampler
name: PKSampler
batch_size: 64
num_instances: 2
sample_per_id: 2
drop_last: False
shuffle: True
loader:
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册