Skip to main content

Table 3 Experiment setting for intra-group priority scheduling

From: A combined priority scheduling method for distributed machine learning

 

Training framework

Number of worker nodes

Number of VM Nodes

Scenarios

Group 1

TensorFlow

5

3

S1

Group 2

TensorFlow

5

3

S2

Group 3

TensorFlow

5

3

S3

Group 4

TensorFlow

5

3

S4

Group 5

MindSpore

5

3

S3

Group 6

TensorFlow

5

6

S3

Group 7

TensorFlow

5

10

S3