pab1 · Sequence CNN

reference_pab1_cnn_plm_contiguousf2_20261104-184720

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	1.3268	0.8232	0.8155	0.6370	1747.0000	0.0526	2802.0000	0.1000	-0.1826	0.1515	0.6049	0.1600	-0.1582	0.2214	0.0000	0.6690	0.2900	-0.1439	0.5640	0.0100	0.7033	0.3000	-0.1316	0.8505	0.0278	0.7348
validation	0.9105	0.8045	0.6946	0.6412	383.0000	-0.0400	1993.0000	0.3000	-0.2372	0.2783	0.4781	0.2200	-0.2900	0.3868	0.0200	0.5192	0.1900	-0.2264	0.4333	0.0700	0.6018	0.1875	-0.2958	0.2783	0.0000	0.4945
test	2.2405	0.7234	0.7578	0.4725	475.0000	-0.0572	1286.0000	0.3000	-0.1584	0.5059	0.7193	0.2200	-0.1621	0.5226	0.0400	0.7717	0.2000	-0.1777	0.5226	0.0400	0.7932	0.2083	-0.1692	0.5226	0.0417	0.7676

Loss per epoch

training validation

Config

name: reference_pab1_cnn_plm_contiguousf2
task: dms
data:
  dataset: pab1
  encoding: one_hot,aa_index,plm,plm_sum
  split_dir: null
  split_scheme: contiguous
  split_fold: 2
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