bgl3 · Fully connected

reference_bgl3_fc_mr5plm_modulof0_20261104-222302

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	1.2930	0.5269	0.5475	0.2190	109.0000	0.0665	821.0000	0.1000	-1.0931	0.1164	0.1659	0.0600	-0.6587	0.3961	0.0400	0.4472	0.0800	-0.7619	0.3961	0.0800	0.4131	0.0690	-0.7886	0.3961	0.0345	0.3506
validation	0.7327	0.5943	0.6112	0.3508	47.0000	-0.1953	126.0000	0.0000	-0.2175	-0.0136	0.5023	0.1600	-0.3156	0.2019	0.1800	0.5805	0.1400	-0.3255	0.2216	0.2800	0.6344	0.0000	-0.1720	-0.0136	0.0000	0.4948
test	0.8561	0.5109	0.5590	0.2586	49.0000	0.1353	21.0000	0.2000	-0.6621	0.2173	0.2631	0.1600	-0.7438	0.7630	0.1600	0.2522	0.1500	-0.6354	0.7630	0.2600	0.3791	0.2857	-0.7076	0.2173	0.0000	0.2264

Loss per epoch

training validation

Config

name: reference_bgl3_fc_mr5plm_modulof0
task: dms
data:
  dataset: bgl3
  encoding: one_hot,aa_index,plm,plm_sum
  split_dir: null
  split_scheme: modulo
  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