gb1 · Sequence CNN

reference_gb1_cnn_plm_contiguousf4_20261105-043315

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	0.2362	0.9790	0.9660	0.9585	30951.0000	0.8720	108.0000	1.0000	1.6575	1.8527	0.8972	1.0000	1.6653	2.0942	0.1200	0.9409	1.0000	1.6721	2.1110	0.2900	0.9612	1.0000	1.3843	2.5295	0.7828	0.9887
validation	2.6844	0.3569	0.3673	0.1255	1866.0000	0.1271	7234.0000	0.7000	-0.1869	0.7502	0.5967	0.6400	-0.1261	0.7876	0.1200	0.6688	0.6200	-0.1505	0.7876	0.1500	0.6792	0.5373	-0.2720	0.9352	0.1343	0.6542
test	5.0540	0.3271	0.2255	-0.1396	5856.0000	0.1364	11497.0000	0.7000	0.0008	0.5617	0.5380	0.4400	-0.3417	0.7608	0.0000	0.4385	0.3500	-0.4948	0.8999	0.0000	0.3928	0.3632	-0.4941	0.8999	0.0000	0.4061

Loss per epoch

training validation

Config

name: reference_gb1_cnn_plm_contiguousf4
task: dms
data:
  dataset: gb1
  encoding: one_hot,aa_index,plm,plm_sum
  split_dir: null
  split_scheme: contiguous
  split_fold: 4
model:
  name: cnn
  conv:
  - - 128
    - 17
  - - 128
    - 17
  - - 128
    - 17
  hidden:
  - 100
  dropout: 0.2
train:
  epochs: 300
  batch_size: 32
  learning_rate: 0.0001
  patience: 15
  min_delta: 1.0e-05
  seed: 7