ube4b · Sequence CNN

reference_ube4b_cnn_plm_contiguousf3_20261105-040915

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	2.5558	0.5883	0.5267	0.3418	5445.0000	0.4201	5777.0000	1.0000	4.3087	5.3352	0.6554	1.0000	4.3863	6.1172	0.1800	0.7731	1.0000	4.2192	6.1172	0.2500	0.8058	0.9955	3.8510	7.2358	0.4136	0.8271
validation	2.3597	0.3907	0.3272	0.1526	933.0000	-0.0644	1091.0000	0.3000	-1.1182	0.3664	-0.0041	0.2400	-0.8993	1.4470	0.0000	0.0476	0.2800	-0.7932	1.5456	0.0000	0.0852	0.2105	-0.9676	0.4566	0.0000	0.0295
test	3.4951	0.3238	0.3465	0.0689	323.0000	-0.0825	489.0000	0.3000	-0.1808	0.7848	0.2584	0.2800	-0.6169	1.4962	0.0000	0.2461	0.3100	-0.4594	3.8792	0.0600	0.3181	0.3600	-0.3833	1.1537	0.0000	0.2585

Loss per epoch

training validation

Config

name: reference_ube4b_cnn_plm_contiguousf3
task: dms
data:
  dataset: ube4b
  encoding: one_hot,aa_index,plm,plm_sum
  split_dir: null
  split_scheme: contiguous
  split_fold: 3
model:
  name: cnn
  conv:
  - - 128
    - 3
  - - 128
    - 3
  - - 128
    - 3
  - - 128
    - 3
  - - 128
    - 3
  hidden:
  - 100
  dropout: 0.2
train:
  epochs: 300
  batch_size: 128
  learning_rate: 0.0001
  patience: 15
  min_delta: 1.0e-05
  seed: 7