add optimizer api docs

[MARD1NO]

增加optimizer文档

warmupLR

optimizer系列

[MARD1NO]

这里的修改是避免下划线让s变成标号形式

[github-actions]

Speed stats:

GPU Name: GeForce GTX 1080 

OneFlow resnet50 time: 128.6ms (= 6428.6ms / 50, input_shape=[16, 3, 224, 224])
PyTorch resnet50 time: 140.8ms (= 7041.0ms / 50, input_shape=[16, 3, 224, 224])
Relative speed: 1.10 (= 140.8ms / 128.6ms)

OneFlow resnet50 time: 74.6ms (= 3728.3ms / 50, input_shape=[8, 3, 224, 224])
PyTorch resnet50 time: 85.1ms (= 4255.5ms / 50, input_shape=[8, 3, 224, 224])
Relative speed: 1.14 (= 85.1ms / 74.6ms)

OneFlow resnet50 time: 50.7ms (= 2533.5ms / 50, input_shape=[4, 3, 224, 224])
PyTorch resnet50 time: 62.6ms (= 3129.9ms / 50, input_shape=[4, 3, 224, 224])
Relative speed: 1.24 (= 62.6ms / 50.7ms)

OneFlow resnet50 time: 44.9ms (= 2245.9ms / 50, input_shape=[2, 3, 224, 224])
PyTorch resnet50 time: 48.7ms (= 2435.9ms / 50, input_shape=[2, 3, 224, 224])
Relative speed: 1.08 (= 48.7ms / 44.9ms)

OneFlow resnet50 time: 40.5ms (= 2026.3ms / 50, input_shape=[1, 3, 224, 224])
PyTorch resnet50 time: 43.3ms (= 2167.3ms / 50, input_shape=[1, 3, 224, 224])
Relative speed: 1.07 (= 43.3ms / 40.5ms)

OneFlow resnet50 time: 146.3ms (= 7314.8ms / 50, input_shape=[16, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 161.1ms (= 8053.8ms / 50, input_shape=[16, 3, 224, 224], ddp, world size=2)
Relative speed: 1.10 (= 161.1ms / 146.3ms)

OneFlow resnet50 time: 94.2ms (= 4709.9ms / 50, input_shape=[8, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 117.8ms (= 5892.4ms / 50, input_shape=[8, 3, 224, 224], ddp, world size=2)
Relative speed: 1.25 (= 117.8ms / 94.2ms)

OneFlow resnet50 time: 71.9ms (= 3595.6ms / 50, input_shape=[4, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 86.4ms (= 4319.9ms / 50, input_shape=[4, 3, 224, 224], ddp, world size=2)
Relative speed: 1.20 (= 86.4ms / 71.9ms)

OneFlow resnet50 time: 57.8ms (= 2888.7ms / 50, input_shape=[2, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 77.7ms (= 3886.8ms / 50, input_shape=[2, 3, 224, 224], ddp, world size=2)
Relative speed: 1.35 (= 77.7ms / 57.8ms)

OneFlow resnet50 time: 65.3ms (= 3264.1ms / 50, input_shape=[1, 3, 224, 224], ddp, world size=2)
PyTorch resnet50 time: 63.2ms (= 3157.6ms / 50, input_shape=[1, 3, 224, 224], ddp, world size=2)
Relative speed: 0.97 (= 63.2ms / 65.3ms)