Step

Train loss

Test loss

0

[2.78e−01, 3.95e−02]

[2.76e−01, 3.95e−02]

1000

[9.28e−05, 4.41e−06]

[8.98e−05, 4.41e−06]

2000

[1.36e−05, 7.99e−07]

[1.32e−05, 7.99e−07]

3000

[7.51e−06, 5.29e−07]

[7.28e−06, 5.29e−07]

4000

[5.12e−06, 1.98e−07]

[4.96e−06, 1.98e−07]

5000

[3.61e−06, 6.31e−08]

[3.50e−06, 6.31e−08]

6000

[2.46e−06, 2.78e−08]

[2.38e−06, 2.27e−08]

7000

[1.70e−06, 6.95e−09]

[1.64e−06, 6.95e−09]

8000

[1.26e−06, 3.64e−10]

[1.21e−06, 3.64e−10]

9000

[1.39e−05, 4.03e−05]

[1.35e−05, 4.03e−05]

10,000

[8.96e−07, 1.43e−08]

[8.49e−07, 1.43e−08]

10,122

[6.43e−08, 1.27e−10]

[6.10e−08, 1.27e−10]