bgl3 · Graph CNN

reference_bgl3_gcn_modulof1_20261104-153356

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	1.7291	0.4748	0.4779	0.0314	3546.0000	0.1187	2198.0000	0.6000	0.3821	2.3637	0.2187	0.4800	0.1796	2.5419	0.0600	0.1820	0.5600	0.2688	2.5419	0.0600	0.2269	0.5556	0.2610	2.5419	0.0606	0.2238
validation	1.5418	-0.0044	-0.0289	-0.0002	432.0000	-0.0270	1300.0000	0.4000	-0.0688	1.6946	0.0000	0.1631	0.4000	-0.7544	1.8690	0.0800	0.0000	-0.0899	0.3400	-0.7353	2.5819	0.0800	0.0000	-0.0903	0.5294	-0.0204	1.6946	0.0000	0.0000	0.1866
test	1.5692	0.1590	0.1476	-0.0024	424.0000	0.0344	52.0000	0.4000	0.1216	1.5945	0.0000	0.1929	0.3400	-0.0653	1.9459	0.0400	0.0000	0.1507	0.4000	-0.0277	2.7932	0.1400	0.0000	0.1867	0.4000	0.2077	1.7856	0.0000	0.0000	0.2297

Loss per epoch

training validation

Config

name: reference_bgl3_gcn_modulof1
task: dms
data:
  dataset: bgl3
  encoding: one_hot,aa_index
  split_dir: null
  graph_threshold: 6
  split_scheme: modulo
  split_fold: 1
model:
  name: gcn
  filters:
  - 32
  hidden:
  - 100
  dropout: 0.2
train:
  epochs: 300
  batch_size: 128
  learning_rate: 0.0001
  patience: 15
  min_delta: 1.0e-05
  seed: 7