Fix BUG of LazyInterpret FreeEagerTensor memory shared with regst

[chengtbf]

• 修复 LazyInterpret 在处理 FreeEagerTensor 时没有绑定好 VariableOpName 的 BUG（错把 lbn 当成了 op name）
• refine 了 测试脚本

[github-actions]

Speed stats:

GPU Name: GeForce GTX 1080 

PyTorch resnet50 time: 141.1ms (= 7053.9ms / 50, input_shape=[16, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 127.8ms (= 6392.0ms / 50, input_shape=[16, 3, 224, 224], backward is enabled)
Relative speed: 1.10 (= 141.1ms / 127.8ms)

PyTorch resnet50 time: 84.2ms (= 4210.6ms / 50, input_shape=[8, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 74.7ms (= 3737.2ms / 50, input_shape=[8, 3, 224, 224], backward is enabled)
Relative speed: 1.13 (= 84.2ms / 74.7ms)

PyTorch resnet50 time: 58.7ms (= 2934.1ms / 50, input_shape=[4, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 47.6ms (= 2379.4ms / 50, input_shape=[4, 3, 224, 224], backward is enabled)
Relative speed: 1.23 (= 58.7ms / 47.6ms)

PyTorch resnet50 time: 49.9ms (= 2495.3ms / 50, input_shape=[2, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 40.2ms (= 2009.1ms / 50, input_shape=[2, 3, 224, 224], backward is enabled)
Relative speed: 1.24 (= 49.9ms / 40.2ms)

PyTorch resnet50 time: 43.5ms (= 2176.6ms / 50, input_shape=[1, 3, 224, 224], backward is enabled)
OneFlow resnet50 time: 41.3ms (= 2064.1ms / 50, input_shape=[1, 3, 224, 224], backward is enabled)
Relative speed: 1.05 (= 43.5ms / 41.3ms)