camillebri commited on
Commit
5645b8f
1 Parent(s): 98cbbb0

Uploading maps_bis in test_maps_bis

Browse files
README.md ADDED
@@ -0,0 +1,89 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ ---
3
+ language: en
4
+ library_name: clinicadl
5
+ tags:
6
+ - clinicadl
7
+ license: mit
8
+ ---
9
+ # Model Card for maps_bis
10
+ This model was trained with ClinicaDL. You can find here all the information.
11
+ ## General information
12
+ This model was trained for **classification** and the architecture chosen is **Conv4_FC3**.
13
+ ### Model
14
+ **architecture**: Conv4_FC3
15
+ **multi_network**: False
16
+ **ssda_network**: False
17
+ ### Architecture
18
+ **dropout**: 0.0
19
+ **latent_space_size**: 2
20
+ **feature_size**: 1024
21
+ **n_conv**: 4
22
+ **io_layer_channels**: 8
23
+ **recons_weight**: 1
24
+ **kl_weight**: 1
25
+ **normalization**: batch
26
+ ### Classification
27
+ **selection_metrics**: ['loss']
28
+ **label**: diagnosis
29
+ **label_code**: {'AD': 0, 'CN': 1}
30
+ **selection_threshold**: 0.0
31
+ **loss**: None
32
+ ### Computational
33
+ **gpu**: True
34
+ **n_proc**: 32
35
+ **batch_size**: 32
36
+ **evaluation_steps**: 20
37
+ **fully_sharded_data_parallel**: False
38
+ **amp**: False
39
+ ### Reproducibility
40
+ **seed**: 0
41
+ **deterministic**: False
42
+ **compensation**: memory
43
+ **track_exp**:
44
+ ### Transfer_learning
45
+ **transfer_path**: ../../autoencoders/exp3/maps
46
+ **transfer_selection_metric**: loss
47
+ **nb_unfrozen_layer**: 0
48
+ ### Mode
49
+ **use_extracted_features**: False
50
+ ### Data
51
+ **multi_cohort**: False
52
+ **diagnoses**: ['AD', 'CN']
53
+ **baseline**: True
54
+ **normalize**: True
55
+ **data_augmentation**: False
56
+ **sampler**: random
57
+ **size_reduction**: False
58
+ **size_reduction_factor**: 2
59
+ **caps_target**:
60
+ **tsv_target_lab**:
61
+ **tsv_target_unlab**:
62
+ **preprocessing_dict_target**:
63
+ ### Cross_validation
64
+ **n_splits**: 5
65
+ **split**: []
66
+ ### Optimization
67
+ **optimizer**: Adam
68
+ **epochs**: 200
69
+ **learning_rate**: 1e-05
70
+ **adaptive_learning_rate**: False
71
+ **weight_decay**: 0.0001
72
+ **patience**: 10
73
+ **tolerance**: 0.0
74
+ **accumulation_steps**: 1
75
+ **profiler**: False
76
+ **save_all_models**: False
77
+ ### Informations
78
+ **emissions_calculator**: False
79
+ ### Other information
80
+ **latent_space_dimension**: 64
81
+ **preprocessing_dict**: {'preprocessing': 't1-linear', 'mode': 'roi', 'use_uncropped_image': False, 'roi_list': ['leftHippocampusBox', 'rightHippocampusBox'], 'uncropped_roi': False, 'prepare_dl': False, 'file_type': {'pattern': '*space-MNI152NLin2009cSym_desc-Crop_res-1x1x1_T1w.nii.gz', 'description': 'T1W Image registered using t1-linear and cropped (matrix size 169×208×179, 1 mm isotropic voxels)', 'needed_pipeline': 't1-linear'}}
82
+ **mode**: roi
83
+ **network_task**: classification
84
+ **caps_directory**: $WORK/../commun/datasets/adni/caps/caps_v2021
85
+ **tsv_path**: $WORK/Aramis_tools/ClinicaDL_tools/experiments_ADDL/data/ADNI/train
86
+ **validation**: KFoldSplit
87
+ **num_networks**: 2
88
+ **output_size**: 2
89
+ **input_size**: [1, 50, 50, 50]
maps.json ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architecture": "Conv4_FC3",
3
+ "multi_network": false,
4
+ "dropout": 0.0,
5
+ "latent_space_dimension": 64,
6
+ "latent_space_size": 2,
7
+ "selection_metrics": [
8
+ "loss"
9
+ ],
10
+ "label": "diagnosis",
11
+ "selection_threshold": 0.0,
12
+ "gpu": true,
13
+ "n_proc": 32,
14
+ "batch_size": 32,
15
+ "evaluation_steps": 20,
16
+ "seed": 0,
17
+ "deterministic": false,
18
+ "compensation": "memory",
19
+ "transfer_path": "../../autoencoders/exp3/maps",
20
+ "transfer_selection_metric": "loss",
21
+ "use_extracted_features": false,
22
+ "multi_cohort": false,
23
+ "diagnoses": [
24
+ "AD",
25
+ "CN"
26
+ ],
27
+ "baseline": true,
28
+ "normalize": true,
29
+ "data_augmentation": false,
30
+ "sampler": "random",
31
+ "n_splits": 5,
32
+ "epochs": 200,
33
+ "learning_rate": 1e-05,
34
+ "weight_decay": 0.0001,
35
+ "patience": 10,
36
+ "tolerance": 0.0,
37
+ "accumulation_steps": 1,
38
+ "optimizer": "Adam",
39
+ "preprocessing_dict": {
40
+ "preprocessing": "t1-linear",
41
+ "mode": "roi",
42
+ "use_uncropped_image": false,
43
+ "roi_list": [
44
+ "leftHippocampusBox",
45
+ "rightHippocampusBox"
46
+ ],
47
+ "uncropped_roi": false,
48
+ "prepare_dl": false,
49
+ "file_type": {
50
+ "pattern": "*space-MNI152NLin2009cSym_desc-Crop_res-1x1x1_T1w.nii.gz",
51
+ "description": "T1W Image registered using t1-linear and cropped (matrix size 169\u00d7208\u00d7179, 1 mm isotropic voxels)",
52
+ "needed_pipeline": "t1-linear"
53
+ }
54
+ },
55
+ "mode": "roi",
56
+ "network_task": "classification",
57
+ "caps_directory": "$WORK/../commun/datasets/adni/caps/caps_v2021/",
58
+ "tsv_path": "$WORK/Aramis_tools/ClinicaDL_tools/experiments_ADDL/data/ADNI/train",
59
+ "validation": "KFoldSplit",
60
+ "num_networks": 2,
61
+ "label_code": {
62
+ "AD": 0,
63
+ "CN": 1
64
+ },
65
+ "output_size": 2,
66
+ "input_size": [
67
+ 1,
68
+ 50,
69
+ 50,
70
+ 50
71
+ ]
72
+ }
split-0/best-loss/model.pth.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35f3385e9e1d0771e5f09ba0fcb5a7f15f4c24732debdf4af2f2478c7606edff
3
+ size 1407722
split-1/best-loss/model.pth.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45f2c49da0cd4f8834b9e78cc8937e503dbd0594b6ed7094f1db3dc030cc7192
3
+ size 1407722
split-2/best-loss/model.pth.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a6ee1ac995daf5604d7efee2d4bbf9312a83c0ea53ef182dbf5f09b52a9bab09
3
+ size 1407722
split-3/best-loss/model.pth.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7df3b31b1f4ae86e72addfdcdd74e67490ce956ce1bd5db476119f7d0c8bd81c
3
+ size 1407722
split-4/best-loss/model.pth.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6827008f14f4f932d8eec1d6af9c5ed7f768e70e9816990fcfa302085c568eb5
3
+ size 1407722