bgl3 · Fully connected

reference_bgl3_fc_mr5_randomf0_20261104-221615

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	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	0.6425	0.8170	0.8179	0.6579	134.0000	0.1535	142.0000	0.3000	-0.0644	0.1391	0.6901	0.2400	-0.1752	0.7042	0.1000	0.7511	0.2200	-0.2165	0.7042	0.1800	0.7774	0.2353	-0.1499	0.7042	0.0882	0.7360
validation	1.0707	0.6613	0.6467	0.4365	37.0000	0.3063	72.0000	0.0000	-0.9456	-0.0549	0.3531	0.1000	-0.8492	0.4856	0.1200	0.4620	0.1000	-0.7523	0.6710	0.2500	0.5501	0.0909	-0.8530	0.0726	0.0000	0.3989
test	1.0702	0.6428	0.6173	0.4106	46.0000	0.1565	184.0000	0.1000	-0.9963	0.1579	0.2836	0.1400	-0.8206	0.7630	0.1600	0.4396	0.0800	-0.8572	0.7630	0.2100	0.4497	0.0909	-0.9285	0.1579	0.0000	0.3192

Loss per epoch

training validation

Config

name: reference_bgl3_fc_mr5_randomf0
task: dms
data:
  dataset: bgl3
  encoding: one_hot,aa_index
  split_dir: null
  split_scheme: random
  split_fold: 0
  min_reads: 5
model:
  name: fcn
  hidden:
  - 100
  - 100
  dropout: 0.2
train:
  epochs: 300
  batch_size: 32
  learning_rate: 0.0001
  patience: 15
  min_delta: 1.0e-05
  seed: 7