pab1 · Sequence CNN

reference_pab1_cnn_plm_modulof4_20261105-042927

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	1.2429	0.8260	0.8133	0.6778	1414.0000	0.0449	3069.0000	0.4000	-0.1247	0.1977	0.6208	0.3400	-0.1191	0.4542	0.0000	0.6910	0.3500	-0.1058	0.5585	0.0200	0.7243	0.3287	-0.1383	0.5585	0.0559	0.7264
validation	1.2457	0.7960	0.7831	0.6122	171.0000	0.0247	297.0000	0.1000	-0.2658	0.0288	0.6185	0.2200	-0.2712	0.2232	0.0200	0.7170	0.2800	-0.2066	0.7072	0.1800	0.7881	0.1765	-0.2272	0.2232	0.0000	0.6700
test	1.2542	0.8245	0.7444	0.6626	234.0000	0.0170	922.0000	0.1000	-0.2910	0.2554	0.6075	0.2200	-0.3420	0.8367	0.0400	0.6426	0.3100	-0.2339	0.8367	0.1400	0.7302	0.1333	-0.2879	0.2554	0.0000	0.6226

Loss per epoch

training validation

Config

name: reference_pab1_cnn_plm_modulof4
task: dms
data:
  dataset: pab1
  encoding: one_hot,aa_index,plm,plm_sum
  split_dir: null
  split_scheme: modulo
  split_fold: 4
model:
  name: cnn
  conv:
  - - 128
    - 17
  - - 128
    - 17
  - - 128
    - 17
  hidden:
  - 100
  dropout: 0.2
train:
  epochs: 300
  batch_size: 128
  learning_rate: 0.0001
  patience: 15
  min_delta: 1.0e-05
  seed: 7