-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathes_original_guo_train.txt
42 lines (42 loc) · 2.1 KB
/
es_original_guo_train.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
[2023-10-27-17:03:35] D-SCRIPT Version 0.2.2
[2023-10-27-17:03:35] Called as: /nfs/home/students/jbernett/miniconda3/envs/dscript/bin/dscript train --train data/original/guo_train_es.txt --test data/original/guo_val_es.txt --embedding /nfs/scratch/jbernett/yeast_embedding.h5 --save-prefix models/es_original_guo -o es_original_guo_train.txt
[2023-10-27-17:03:35] Loaded 12618 training pairs
[2023-10-27-17:03:35] Loaded 701 test pairs
[2023-10-27-17:03:35] Loading embeddings...
[2023-10-27-17:04:07] Initializing embedding model with:
[2023-10-27-17:04:07] projection_dim: 100
[2023-10-27-17:04:07] dropout_p: 0.5
[2023-10-27-17:04:07] Initializing contact model with:
[2023-10-27-17:04:07] hidden_dim: 50
[2023-10-27-17:04:07] kernel_width: 7
[2023-10-27-17:04:07] Initializing interaction model with:
[2023-10-27-17:04:07] do_poool: False
[2023-10-27-17:04:07] pool_width: 9
[2023-10-27-17:04:07] do_w: True
[2023-10-27-17:04:07] do_sigmoid: True
[2023-10-27-17:04:07] ModelInteraction(
(activation): LogisticActivation()
(embedding): FullyConnectedEmbed(
(transform): Linear(in_features=6165, out_features=100, bias=True)
(drop): Dropout(p=0.5, inplace=False)
(activation): ReLU()
)
(contact): ContactCNN(
(hidden): FullyConnected(
(conv): Conv2d(200, 50, kernel_size=(1, 1), stride=(1, 1))
(batchnorm): BatchNorm2d(50, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(activation): ReLU()
)
(conv): Conv2d(50, 1, kernel_size=(7, 7), stride=(1, 1), padding=(3, 3))
(batchnorm): BatchNorm2d(1, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(activation): Sigmoid()
)
(maxPool): MaxPool2d(kernel_size=9, stride=9, padding=4, dilation=1, ceil_mode=False)
)
[2023-10-27-17:04:08] Using save prefix "models/es_original_guo"
[2023-10-27-17:04:08] Training with Adam: lr=0.001, weight_decay=0
[2023-10-27-17:04:08] num_epochs: 10
[2023-10-27-17:04:08] batch_size: 25
[2023-10-27-17:04:08] interaction weight: 0.35
[2023-10-27-17:04:08] contact map weight: 0.65
[2023-10-27-17:04:15] [1/10] training 0.8%: Loss=1.43536, Accuracy=50.000%, MSE=0.49819