gb1 · Sequence CNN

reference_gb1_cnn_modulof1_20261104-153415

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	0.3311	0.9718	0.9721	0.9354	35035.0000	0.8378	195.0000	1.0000	1.6062	1.9240	0.8586	1.0000	1.6141	2.0942	0.0600	0.9104	1.0000	1.6075	2.1952	0.2200	0.9322	0.9995	1.3718	2.5295	0.7096	0.9797
validation	9.1854	-0.2723	-0.3019	-1.0686	1545.0000	0.0351	12820.0000	0.0000	-4.9819	-3.0400	0.0000	-0.5921	0.0000	-4.6529	-2.5067	0.0000	0.0000	-0.5254	0.0000	-4.6650	-2.1718	0.0000	0.0000	-0.5392	0.0000	-4.3877	-0.9853	0.0000	0.0000	-0.4706
test	6.4741	0.0846	0.0597	-0.1165	4408.0000	-0.0002	10010.0000	0.0000	-1.2419	-0.3130	0.0000	0.2882	0.1000	-2.0838	0.7449	0.0000	0.0000	0.0758	0.1000	-2.0248	0.8670	0.0000	0.0000	0.0941	0.0804	-2.1134	0.9059	0.0000	0.0000	0.0713

Loss per epoch

training validation

Config

name: reference_gb1_cnn_modulof1
task: dms
data:
  dataset: gb1
  encoding: one_hot,aa_index
  split_dir: null
  split_scheme: modulo
  split_fold: 1
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