Skip to content

Commit 8e8e81e

Browse files
melvinbarbauxmelvinbarbaux
authored
feat(configs): add best experiment YAMLs for label fractions p01/p02/p05/p10 (#34)
1 parent d8f1b0a commit 8e8e81e

File tree

6,360 files changed

+526307
-102924
lines changed

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

6,360 files changed

+526307
-102924
lines changed

‎bench/configs/experiments/best/README_by_dataset.md‎

Lines changed: 0 additions & 1151 deletions
This file was deleted.

‎bench/configs/experiments/best/README_by_method.md‎

Lines changed: 0 additions & 1358 deletions
This file was deleted.

‎bench/configs/experiments/best/commands/01_by_dataset.md‎

Lines changed: 25 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,25 @@
1+
# Raw Commands By Dataset
2+
3+
Raw commands are split into smaller files:
4+
5+
- `by_dataset/adult.md`
6+
- `by_dataset/ag_news.md`
7+
- `by_dataset/amazon_polarity.md`
8+
- `by_dataset/amazon_reviews_multi_en.md`
9+
- `by_dataset/breast_cancer.md`
10+
- `by_dataset/cifar10.md`
11+
- `by_dataset/cifar100.md`
12+
- `by_dataset/citeseer.md`
13+
- `by_dataset/cora.md`
14+
- `by_dataset/dbpedia_14.md`
15+
- `by_dataset/imdb.md`
16+
- `by_dataset/iris.md`
17+
- `by_dataset/mnist.md`
18+
- `by_dataset/pubmed.md`
19+
- `by_dataset/speechcommands.md`
20+
- `by_dataset/stl10.md`
21+
- `by_dataset/svhn.md`
22+
- `by_dataset/toy.md`
23+
- `by_dataset/yelp_polarity.md`
24+
- `by_dataset/yelp_review_full.md`
25+
- `by_dataset/yesno.md`

‎bench/configs/experiments/best/commands/02_by_method.md‎

Lines changed: 55 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,55 @@
1+
# Raw Commands By Method
2+
3+
Raw commands are split into smaller files:
4+
5+
- `by_method/adamatch.md`
6+
- `by_method/adsh.md`
7+
- `by_method/appnp.md`
8+
- `by_method/chebnet.md`
9+
- `by_method/co_training.md`
10+
- `by_method/comatch.md`
11+
- `by_method/daso.md`
12+
- `by_method/deep_co_training.md`
13+
- `by_method/defixmatch.md`
14+
- `by_method/democratic_co_learning.md`
15+
- `by_method/dynamic_label_propagation.md`
16+
- `by_method/fixmatch.md`
17+
- `by_method/flexmatch.md`
18+
- `by_method/free_match.md`
19+
- `by_method/gat.md`
20+
- `by_method/gcn.md`
21+
- `by_method/gcnii.md`
22+
- `by_method/grafn.md`
23+
- `by_method/grand.md`
24+
- `by_method/graph_mincuts.md`
25+
- `by_method/graphhop.md`
26+
- `by_method/graphsage.md`
27+
- `by_method/h_gcn.md`
28+
- `by_method/label_propagation.md`
29+
- `by_method/label_spreading.md`
30+
- `by_method/laplace_learning.md`
31+
- `by_method/lazy_random_walk.md`
32+
- `by_method/mean_teacher.md`
33+
- `by_method/meta_pseudo_labels.md`
34+
- `by_method/mixmatch.md`
35+
- `by_method/n_gcn.md`
36+
- `by_method/noisy_student.md`
37+
- `by_method/p_laplace_learning.md`
38+
- `by_method/pi_model.md`
39+
- `by_method/planetoid.md`
40+
- `by_method/poisson_learning.md`
41+
- `by_method/poisson_mbo.md`
42+
- `by_method/pseudo_label.md`
43+
- `by_method/s4vm.md`
44+
- `by_method/self_training.md`
45+
- `by_method/setred.md`
46+
- `by_method/sgc.md`
47+
- `by_method/simclr_v2.md`
48+
- `by_method/softmatch.md`
49+
- `by_method/supervised.md`
50+
- `by_method/temporal_ensembling.md`
51+
- `by_method/tri_training.md`
52+
- `by_method/trinet.md`
53+
- `by_method/tsvm.md`
54+
- `by_method/uda.md`
55+
- `by_method/vat.md`

‎bench/configs/experiments/best/commands/03_by_modality.md‎

Lines changed: 9 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,9 @@
1+
# Raw Commands By Modality
2+
3+
Raw commands are split into smaller files:
4+
5+
- `by_modality/audio.md`
6+
- `by_modality/graph.md`
7+
- `by_modality/tabular.md`
8+
- `by_modality/text.md`
9+
- `by_modality/vision.md`

‎bench/configs/experiments/best/commands/04_by_percentage.md‎

Lines changed: 9 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,9 @@
1+
# Raw Commands By Label Percentage
2+
3+
Raw commands are split into smaller files:
4+
5+
- `by_percentage/p01.md`
6+
- `by_percentage/p02.md`
7+
- `by_percentage/p05.md`
8+
- `by_percentage/p10.md`
9+
- `by_percentage/p20.md`

‎bench/configs/experiments/best/commands/05_lightest_per_modality.md‎

Lines changed: 1369 additions & 0 deletions
Large diffs are not rendered by default.

‎bench/configs/experiments/best/commands/README.md‎

Lines changed: 33 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,33 @@
1+
# Best Sweep Command Runbooks
2+
3+
All command files are in English and contain raw `python -m bench.main --config ...` lines.
4+
5+
## Label Percentages
6+
7+
- `p01`
8+
- `p02`
9+
- `p05`
10+
- `p10`
11+
- `p20`
12+
13+
## Modalities And Datasets
14+
15+
- `audio`: `speechcommands, yesno`
16+
- `graph`: `citeseer, cora, pubmed`
17+
- `tabular`: `adult, breast_cancer, iris, toy`
18+
- `text`: `ag_news, amazon_polarity, amazon_reviews_multi_en, dbpedia_14, imdb, yelp_polarity, yelp_review_full`
19+
- `vision`: `cifar10, cifar100, mnist, stl10, svhn`
20+
21+
## Index Files
22+
23+
- `01_by_dataset.md` -> split files in `by_dataset/`
24+
- `02_by_method.md` -> split files in `by_method/`
25+
- `03_by_modality.md` -> split files in `by_modality/`
26+
- `04_by_percentage.md` -> split files in `by_percentage/`
27+
- `05_lightest_per_modality.md` -> all methods x all percentages using one light dataset per modality
28+
29+
## Command Format
30+
31+
```bash
32+
python -m bench.main --config bench/configs/experiments/best/<pXX>/<kind>/<method>/<modality>/<dataset>.yaml
33+
```

‎bench/configs/experiments/best/commands/by_dataset/adult.md‎

Lines changed: 281 additions & 0 deletions
Large diffs are not rendered by default.

‎bench/configs/experiments/best/commands/by_dataset/ag_news.md‎

Lines changed: 266 additions & 0 deletions
Large diffs are not rendered by default.

0 commit comments

Comments
 (0)