bgl3 · Fully connected

reference_bgl3_fc_modulof1_20261104-153347

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	1.8163	0.5075	0.5026	-0.0174	3546.0000	0.1509	2601.0000	0.7000	0.4809	2.0174	0.2480	0.7000	0.6021	2.8245	0.0200	0.3262	0.7100	0.6585	2.8245	0.0700	0.3689	0.7071	0.6558	2.8245	0.0707	0.3675
validation	1.5429	0.0120	0.0177	-0.0009	432.0000	0.0767	1372.0000	0.1000	-0.6177	1.8690	0.0000	-0.0295	0.2400	-0.4929	1.8690	0.0400	0.0000	0.0166	0.2900	-0.5771	2.0174	0.0700	0.0000	-0.0194	0.2353	-0.2533	1.8690	0.0000	0.0000	0.1019
test	1.5836	0.0095	0.0173	-0.0116	424.0000	0.0549	812.0000	0.2000	-0.3542	1.2150	0.0000	0.0287	0.3400	-0.3168	1.8690	0.0200	0.0000	0.0488	0.4200	-0.1442	2.6574	0.1100	0.0000	0.1336	0.3333	-0.2252	1.8690	0.0000	0.0000	0.0756

Loss per epoch

training validation

Config

name: reference_bgl3_fc_modulof1
task: dms
data:
  dataset: bgl3
  encoding: one_hot,aa_index
  split_dir: null
  split_scheme: modulo
  split_fold: 1
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