pab1 · Fully connected

reference_pab1_fc_modulof1_20261104-153753

Metrics

Set	mse	pearsonr	spearmanr	r2	above_wt_n	above_wt_spearmanr	budget_to_max	hit_rate_at_10	mean_at_10	max_at_10	precision_at_10	captured_at_10	hit_rate_at_50	mean_at_50	max_at_50	precision_at_50	captured_at_50	hit_rate_at_100	mean_at_100	max_at_100	precision_at_100	captured_at_100	hit_rate_at_1pct	mean_at_1pct	max_at_1pct	precision_at_1pct	captured_at_1pct
train	2.2104	0.7293	0.6860	0.3847	1937.0000	0.0942	7287.0000	0.1000	-0.5107	0.0967	0.0000	0.4514	0.2000	-0.2537	0.8908	0.0200	0.5960	0.1600	-0.2582	0.8908	0.0300	0.6212	0.1812	-0.3468	0.8908	0.0336	0.6001
validation	3.6472	0.0494	0.0378	-0.0160	116.0000	0.0814	2.0000	0.4000	-0.7528	0.7547	0.1000	0.0000	0.4811	0.3200	-0.7035	0.7547	0.1800	0.0000	0.5512	0.2000	-1.4498	0.7547	0.1600	0.0000	0.2270	0.2857	-0.7527	0.7547	0.0714	0.0000	0.4899
test	4.0617	-0.2331	-0.1960	-0.0175	132.0000	0.1284	1612.0000	0.2000	-2.8714	0.2996	0.0000	0.0000	-0.2230	0.1200	-3.0528	0.2996	0.0600	0.0000	-0.3232	0.0700	-3.5470	0.2996	0.0400	0.0000	-0.5418	0.1765	-2.4576	0.2996	0.0588	0.0000	-0.0815

Loss per epoch

training validation

Config

name: reference_pab1_fc_modulof1
task: dms
data:
  dataset: pab1
  encoding: one_hot,aa_index
  split_dir: null
  split_scheme: modulo
  split_fold: 1
model:
  name: fcn
  hidden:
  - 100
  - 100
  - 100
  dropout: 0.2
train:
  epochs: 300
  batch_size: 128
  learning_rate: 0.001
  patience: 15
  min_delta: 1.0e-05
  seed: 7