bgl3 · Sequence CNN

reference_bgl3_cnn_mr5plm_modulof1_20261104-222448

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	0.5524	0.8204	0.8247	0.6524	136.0000	0.0374	829.0000	0.5000	0.0654	0.3887	0.7281	0.3200	-0.0798	0.5578	0.1400	0.7537	0.3100	-0.0924	0.5578	0.2300	0.8035	0.2963	-0.0784	0.5578	0.1111	0.7087
validation	0.8220	0.5550	0.6271	0.3003	23.0000	-0.1028	23.0000	0.0000	-0.7214	-0.0033	0.2799	0.1000	-0.4037	0.3940	0.2000	0.6054	0.1500	-0.3547	0.3940	0.4000	0.6990	0.0000	-0.6805	-0.0033	0.0000	0.3032
test	0.7073	0.6181	0.6344	0.3733	47.0000	0.0860	167.0000	0.3000	-0.1320	0.1910	0.5587	0.2800	-0.2189	0.3141	0.2800	0.6652	0.1800	-0.2880	0.3141	0.3400	0.6709	0.2857	-0.1421	0.1579	0.0000	0.5131

Loss per epoch

training validation

Config

name: reference_bgl3_cnn_mr5plm_modulof1
task: dms
data:
  dataset: bgl3
  encoding: one_hot,aa_index,plm,plm_sum
  split_dir: null
  split_scheme: modulo
  split_fold: 1
  min_reads: 5
model:
  name: cnn
  conv:
  - - 32
    - 17
  hidden:
  - 100
  dropout: 0.2
train:
  epochs: 300
  batch_size: 64
  learning_rate: 0.0001
  patience: 15
  min_delta: 1.0e-05
  seed: 7