bgl3 · Fully connected

reference_bgl3_fc_plm_modulof2_20261104-183542

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	1.6955	0.3075	0.2546	0.0725	3477.0000	-0.0657	4476.0000	0.6000	0.7530	2.1466	0.3392	0.5000	0.2735	2.2054	0.0000	0.2149	0.5300	0.2617	2.2054	0.0300	0.2232	0.5263	0.2407	2.2054	0.0211	0.2145
validation	1.2868	0.2631	0.2272	0.0667	437.0000	-0.1218	481.0000	0.3000	-0.2668	2.2054	0.0474	0.2600	-0.2683	2.2054	0.0600	0.0590	0.3000	-0.2382	2.2391	0.0700	0.0808	0.3125	-0.0391	2.2054	0.0625	0.1275
test	1.4106	0.2948	0.2984	0.0849	432.0000	-0.0650	646.0000	0.4000	-0.8504	1.3581	-0.1111	0.3400	-0.3927	1.5945	0.0200	0.0574	0.3600	-0.2807	2.3637	0.1000	0.1133	0.4118	-0.5424	1.5945	0.0000	-0.0031

Loss per epoch

training validation

Config

name: reference_bgl3_fc_plm_modulof2
task: dms
data:
  dataset: bgl3
  encoding: one_hot,aa_index,plm,plm_sum
  split_dir: null
  split_scheme: modulo
  split_fold: 2
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