learning rate: 0.01618:32, 0.01:64, 0.00618:192, 0.003819:400, 0.00236:800, 0.00146:1024, 0.000902:1280,0.0005
Batch size :1
Gradient accumulation steps: 1
learning rate: 0.01618:32, 0.01:64, 0.00618:192, 0.003819:400, 0.00236:800, 0.00146:1024, 0.000902:1280,0.0005
Batch size :1
Gradient accumulation steps: 1