|
@ -1,11 +1,14 @@ |
|
|
import subprocess |
|
|
import subprocess |
|
|
import sys |
|
|
import sys |
|
|
from random import sample |
|
|
from random import sample, seed |
|
|
|
|
|
|
|
|
import numpy as np # noqa: F401 |
|
|
import numpy as np # noqa: F401 |
|
|
from lightning_sdk import Machine, Studio # noqa: F401 |
|
|
from lightning_sdk import Machine, Studio # noqa: F401 |
|
|
|
|
|
|
|
|
NUM_JOBS = 500 |
|
|
# consistency of randomly sampled experiments. |
|
|
|
|
|
seed(19920921) |
|
|
|
|
|
|
|
|
|
|
|
NUM_JOBS = 100 |
|
|
|
|
|
|
|
|
# reference to the current studio |
|
|
# reference to the current studio |
|
|
# if you run outside of Lightning, you can pass the Studio name |
|
|
# if you run outside of Lightning, you can pass the Studio name |
|
@ -22,10 +25,10 @@ NUM_JOBS = 500 |
|
|
# learning_rate_values = list(np.round(np.logspace(-5, -3, 21), 5)) |
|
|
# learning_rate_values = list(np.round(np.logspace(-5, -3, 21), 5)) |
|
|
learning_rate_values = [1e-2] |
|
|
learning_rate_values = [1e-2] |
|
|
alpha_values = [0] |
|
|
alpha_values = [0] |
|
|
widths = [2**k for k in range(4, 15)] |
|
|
widths = [2**k for k in range(4, 13)] |
|
|
depths = [1, 2, 4, 8, 16] |
|
|
depths = [1, 2, 4, 8, 16] |
|
|
# learning_rate_values = [5e-4] |
|
|
# learning_rate_values = [5e-4] |
|
|
batch_size_values = [256] |
|
|
batch_size_values = [16, 64, 256] |
|
|
max_epochs_values = [100] |
|
|
max_epochs_values = [100] |
|
|
seeds = list(range(21, 1992)) |
|
|
seeds = list(range(21, 1992)) |
|
|
optimizers = [ |
|
|
optimizers = [ |
|
@ -67,6 +70,7 @@ python newmain.py fit \ |
|
|
--model.alpha {a} \ |
|
|
--model.alpha {a} \ |
|
|
--model.width {w} \ |
|
|
--model.width {w} \ |
|
|
--model.depth {d} \ |
|
|
--model.depth {d} \ |
|
|
|
|
|
--model.bias true \ |
|
|
--trainer.min_epochs 10 \ |
|
|
--trainer.min_epochs 10 \ |
|
|
--trainer.max_epochs {me} \ |
|
|
--trainer.max_epochs {me} \ |
|
|
--trainer.log_every_n_steps 3 \ |
|
|
--trainer.log_every_n_steps 3 \ |
|
|