Skip to content

chore: stable transductive methods benchmark release 0.1.0 #23

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
melvinbarbaux merged 2 commits into from
Jan 12, 2026
Merged

chore: stable transductive methods benchmark release 0.1.0 #23

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
757e822
chore: stable transductive methods benchmark release 0.1.0
melvinbarbaux Jan 12, 2026
e87adc1
fix(transductive): infer class count from full labels and update tests
melvinbarbaux Jan 12, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
3 changes: 3 additions & 0 deletions CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,9 @@ All notable changes to this project will be documented in this file.

The format is based on "Keep a Changelog", and this project adheres to Semantic Versioning.

## 0.1.0
- Contains all desired transductive methods and marks a stable benchmark release.

## 0.0.4
- Added multiple semi-supervised methods.

Expand Down
1,390 changes: 837 additions & 553 deletions bench/configs/experiments/best/README_by_dataset.md
View file
Open in desktop

Large diffs are not rendered by default.

405 changes: 405 additions & 0 deletions bench/configs/experiments/best/README_by_method.md
View file
Open in desktop

Large diffs are not rendered by default.

97 changes: 97 additions & 0 deletions bench/configs/experiments/best/inductive/adsh/audio/speechcommands.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,97 @@
run:
name: best_audio_inductive_adsh_speechcommands
seed: 4
output_dir: runs/inductive/adsh/audio/speechcommands
fail_fast: true
log_level: detailed

limits:
profile: auto

dataset:
id: speechcommands
download: true
sampling:
seed: 4
plan:
split:
kind: holdout
test_fraction: 0.2
val_fraction: 0.1
stratify: true
shuffle: true
labeling:
mode: fraction
value: 0.2
strategy: balanced
min_per_class: 1
imbalance:
kind: none
policy:
respect_official_test: true
use_official_graph_masks: true
allow_override_official: false
preprocess:
seed: 4
fit_on: train_labeled
cache: true
plan:
output_key: features.X
steps:
- id: labels.encode
- id: audio.load_waveform
params:
target_sample_rate: 16000
max_length: 16000
- id: core.to_torch
params:
device: "auto"
dtype: float32
augmentation:
enabled: true
seed: 4
mode: fixed
modality: tabular
weak:
steps:
- id: tabular.gaussian_noise
params:
std: 0.01
strong:
steps:
- id: tabular.feature_dropout
params:
p: 0.2
method:
kind: inductive
id: adsh
device:
device: "auto"
dtype: float32
params:
lambda_u: 1.0
p_cutoff: 0.95
use_cat: false
batch_size: 128
max_epochs: 50
detach_target: true
score_warmup_epochs: 2
model:
classifier_id: audio_pretrained
classifier_backend: torch
classifier_params:
bundle: WAV2VEC2_BASE
freeze_backbone: true
lr: 0.0001
weight_decay: 0.0001
batch_size: 8
max_epochs: 5
ema: false
evaluation:
split_for_model_selection: val
report_splits:
- val
- test
metrics:
- accuracy
- macro_f1
97 changes: 97 additions & 0 deletions bench/configs/experiments/best/inductive/adsh/audio/yesno.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,97 @@
run:
name: best_audio_inductive_adsh_yesno
seed: 4
output_dir: runs/inductive/adsh/audio/yesno
fail_fast: true
log_level: detailed

limits:
profile: auto

dataset:
id: yesno
download: true
sampling:
seed: 4
plan:
split:
kind: holdout
test_fraction: 0.2
val_fraction: 0.1
stratify: true
shuffle: true
labeling:
mode: fraction
value: 0.2
strategy: balanced
min_per_class: 1
imbalance:
kind: none
policy:
respect_official_test: true
use_official_graph_masks: true
allow_override_official: false
preprocess:
seed: 4
fit_on: train_labeled
cache: true
plan:
output_key: features.X
steps:
- id: labels.encode
- id: audio.load_waveform
params:
target_sample_rate: 16000
max_length: 16000
- id: core.to_torch
params:
device: "auto"
dtype: float32
augmentation:
enabled: true
seed: 4
mode: fixed
modality: tabular
weak:
steps:
- id: tabular.gaussian_noise
params:
std: 0.01
strong:
steps:
- id: tabular.feature_dropout
params:
p: 0.2
method:
kind: inductive
id: adsh
device:
device: "auto"
dtype: float32
params:
lambda_u: 1.0
p_cutoff: 0.95
use_cat: false
batch_size: 128
max_epochs: 50
detach_target: true
score_warmup_epochs: 2
model:
classifier_id: audio_pretrained
classifier_backend: torch
classifier_params:
bundle: WAV2VEC2_BASE
freeze_backbone: true
lr: 0.0001
weight_decay: 0.0001
batch_size: 8
max_epochs: 5
ema: false
evaluation:
split_for_model_selection: val
report_splits:
- val
- test
metrics:
- accuracy
- macro_f1
109 changes: 109 additions & 0 deletions bench/configs/experiments/best/inductive/adsh/graph/citeseer.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,109 @@
run:
name: best_graph_inductive_adsh_citeseer
seed: 5
output_dir: runs/inductive/adsh/graph/citeseer
fail_fast: true
log_level: detailed

limits:
profile: auto

dataset:
id: citeseer
download: true
sampling:
seed: 5
plan:
split:
kind: holdout
test_fraction: 0.2
val_fraction: 0.1
stratify: true
shuffle: true
labeling:
mode: fraction
value: 0.2
strategy: balanced
min_per_class: 1
imbalance:
kind: none
policy:
respect_official_test: true
use_official_graph_masks: true
allow_override_official: false
preprocess:
seed: 5
fit_on: train_labeled
cache: true
plan:
output_key: features.X
steps:
- id: labels.encode
- id: graph.node2vec
params:
embedding_dim: 128
num_walks: 10
walk_length: 40
window_size: 5
p: 1.0
q: 1.0
num_negative: 5
batch_size: 1024
embed_epochs: 5
embed_lr: 0.01
undirected: true
device: "auto"
- id: core.to_torch
params:
device: "auto"
dtype: float32
augmentation:
enabled: true
seed: 5
mode: fixed
modality: tabular
weak:
steps:
- id: tabular.gaussian_noise
params:
std: 0.01
strong:
steps:
- id: tabular.feature_dropout
params:
p: 0.2
method:
kind: inductive
id: adsh
device:
device: "auto"
dtype: float32
params:
lambda_u: 1.0
p_cutoff: 0.95
use_cat: false
batch_size: 128
max_epochs: 50
detach_target: true
score_warmup_epochs: 2
model:
classifier_id: mlp
classifier_backend: torch
classifier_params:
hidden_sizes:
- 128
activation: relu
dropout: 0.1
lr: 0.001
weight_decay: 0.0
batch_size: 256
max_epochs: 50
ema: false
evaluation:
split_for_model_selection: val
report_splits:
- val
- test
metrics:
- accuracy
- macro_f1
Loading
Loading