Skip to content

Forward-merge branch-25.08 into branch-25.10 #7103

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
rapids-bot merged 3 commits into from
Aug 11, 2025
Merged

Forward-merge branch-25.08 into branch-25.10 #7103

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
34dabf8
Fix network failure in loading mnist dataset during spectral embeddin…
aamijar Aug 4, 2025
00094f7
Update Changelog [skip ci]
AyodeAwe Aug 6, 2025
472cacb
Merge branch 'branch-25.08' into 'branch-25.10'
jcrist Aug 8, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
143 changes: 143 additions & 0 deletions CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,3 +1,146 @@
# cuml 25.08.00 (6 Aug 2025)

## 🚨 Breaking Changes

- Deprecating `data_on_host` parameter for UMAP ([#6953](https://github.com/rapidsai/cuml/pull/6953)) [@jinsolp](https://github.com/jinsolp)
- Deprecate hdbscan prediction functions in `cuml.cluster` namespace ([#6943](https://github.com/rapidsai/cuml/pull/6943)) [@jcrist](https://github.com/jcrist)
- Reflect output type in `hdbscan` prediction functions ([#6939](https://github.com/rapidsai/cuml/pull/6939)) [@jcrist](https://github.com/jcrist)
- Deprecate `connectivity` parameter to `HDBSCAN` ([#6936](https://github.com/rapidsai/cuml/pull/6936)) [@jcrist](https://github.com/jcrist)
- Deprecate `penalty='none'` in `MBSGDClassifier`/`MBSGDRegressor`/`SGD` ([#6926](https://github.com/rapidsai/cuml/pull/6926)) [@jcrist](https://github.com/jcrist)
- Remove `experimental.fil` Python module ([#6899](https://github.com/rapidsai/cuml/pull/6899)) [@hcho3](https://github.com/hcho3)
- Change `KMeans` `random_state` default to `None` ([#6884](https://github.com/rapidsai/cuml/pull/6884)) [@jcrist](https://github.com/jcrist)
- Remove cuvs symbols in libcuml headers ([#6883](https://github.com/rapidsai/cuml/pull/6883)) [@hcho3](https://github.com/hcho3)
- Migrate C++ files from `src/experimental/fil` to `src/fil` ([#6882](https://github.com/rapidsai/cuml/pull/6882)) [@hcho3](https://github.com/hcho3)
- Add test for signatures of common methods ([#6878](https://github.com/rapidsai/cuml/pull/6878)) [@jcrist](https://github.com/jcrist)
- Port `cuml.neighbors` to `InteropMixin`/`ProxyBase` ([#6851](https://github.com/rapidsai/cuml/pull/6851)) [@jcrist](https://github.com/jcrist)
- Remove CUDA 11 from dependencies.yaml ([#6847](https://github.com/rapidsai/cuml/pull/6847)) [@KyleFromNVIDIA](https://github.com/KyleFromNVIDIA)
- Remove legacy FIL from libcuml ([#6844](https://github.com/rapidsai/cuml/pull/6844)) [@hcho3](https://github.com/hcho3)
- stop uploading packages to downloads.rapids.ai ([#6803](https://github.com/rapidsai/cuml/pull/6803)) [@jameslamb](https://github.com/jameslamb)

## 🐛 Bug Fixes

- Fix network failure in loading mnist dataset during spectral embedding testing ([#7077](https://github.com/rapidsai/cuml/pull/7077)) [@aamijar](https://github.com/aamijar)
- Relaxing unit tolerance for `test_tsne_distance_metrics_on_sparse_input` ([#7033](https://github.com/rapidsai/cuml/pull/7033)) [@jinsolp](https://github.com/jinsolp)
- Disable UMAP non-determinism on small datasets ([#7004](https://github.com/rapidsai/cuml/pull/7004)) [@viclafargue](https://github.com/viclafargue)
- Remove CuPy version constraint ([#6979](https://github.com/rapidsai/cuml/pull/6979)) [@csadorf](https://github.com/csadorf)
- Fix test_typeerror_input test with xfail for known sklearn LARS broadcasting bug ([#6978](https://github.com/rapidsai/cuml/pull/6978)) [@csadorf](https://github.com/csadorf)
- Add upper constraint for cupy 13.5 due to memory access issues ([#6961](https://github.com/rapidsai/cuml/pull/6961)) [@csadorf](https://github.com/csadorf)
- Decrease nightly check window to 7 days ([#6934](https://github.com/rapidsai/cuml/pull/6934)) [@csadorf](https://github.com/csadorf)
- Support empty categorical nodes in FIL ([#6924](https://github.com/rapidsai/cuml/pull/6924)) [@hcho3](https://github.com/hcho3)
- Rewrite `HDBSCAN` python wrapper ([#6913](https://github.com/rapidsai/cuml/pull/6913)) [@jcrist](https://github.com/jcrist)
- Mark test_warm_start_oob[RandomForestClassifier] as flaky ([#6906](https://github.com/rapidsai/cuml/pull/6906)) [@csadorf](https://github.com/csadorf)
- Unxfail tests affected by numba compilation error on Python 3.13.4 ([#6905](https://github.com/rapidsai/cuml/pull/6905)) [@csadorf](https://github.com/csadorf)
- Fix import in benchmark code in algorithms.py that was causing an error ([#6902](https://github.com/rapidsai/cuml/pull/6902)) [@dantegd](https://github.com/dantegd)
- Adding GH_TOKEN pass-through to summarize job ([#6894](https://github.com/rapidsai/cuml/pull/6894)) [@msarahan](https://github.com/msarahan)
- Add test for signatures of common methods ([#6878](https://github.com/rapidsai/cuml/pull/6878)) [@jcrist](https://github.com/jcrist)
- Fix CI test failures: xfail tests for scikit-learn 1.7.0 and Python 3.13.4 compatibility ([#6865](https://github.com/rapidsai/cuml/pull/6865)) [@csadorf](https://github.com/csadorf)
- Fix RandomForestRegressor default `max_features` ([#6862](https://github.com/rapidsai/cuml/pull/6862)) [@jcrist](https://github.com/jcrist)
- Increase tolerance in Dask logistic regression test ([#6848](https://github.com/rapidsai/cuml/pull/6848)) [@csadorf](https://github.com/csadorf)

## 📖 Documentation

- Document supported versions. ([#7040](https://github.com/rapidsai/cuml/pull/7040)) [@csadorf](https://github.com/csadorf)
- Update title of `cuml.accel` documentation to "Zero Code Change Acceleration" ([#7030](https://github.com/rapidsai/cuml/pull/7030)) [@csadorf](https://github.com/csadorf)
- Reorganize cuml.accel docs ([#7026](https://github.com/rapidsai/cuml/pull/7026)) [@jcrist](https://github.com/jcrist)
- UMAP multi-gpu knn graph build docs ([#7019](https://github.com/rapidsai/cuml/pull/7019)) [@jinsolp](https://github.com/jinsolp)
- Remove async operation section from Python developer docs ([#6980](https://github.com/rapidsai/cuml/pull/6980)) [@csadorf](https://github.com/csadorf)
- Fix docs for FFT TSNE ([#6967](https://github.com/rapidsai/cuml/pull/6967)) [@jinsolp](https://github.com/jinsolp)
- Revamp `cuml.accel` limitations docs page ([#6965](https://github.com/rapidsai/cuml/pull/6965)) [@jcrist](https://github.com/jcrist)
- add docs on CI workflow inputs ([#6952](https://github.com/rapidsai/cuml/pull/6952)) [@jameslamb](https://github.com/jameslamb)
- Fixup rst formatting of 0cc limitations doc ([#6945](https://github.com/rapidsai/cuml/pull/6945)) [@jcrist](https://github.com/jcrist)
- Documentation Updates for cuML Python Developer Guide ([#6843](https://github.com/rapidsai/cuml/pull/6843)) [@csadorf](https://github.com/csadorf)

## 🚀 New Features

- Spectral Embedding ([#6581](https://github.com/rapidsai/cuml/pull/6581)) [@aamijar](https://github.com/aamijar)

## 🛠️ Improvements

- Add profilers to `cuml.accel` ([#7021](https://github.com/rapidsai/cuml/pull/7021)) [@jcrist](https://github.com/jcrist)
- use versioned CI image tags ([#7016](https://github.com/rapidsai/cuml/pull/7016)) [@jameslamb](https://github.com/jameslamb)
- Support inferencing with wide data in FIL ([#7014](https://github.com/rapidsai/cuml/pull/7014)) [@hcho3](https://github.com/hcho3)
- Remove sphinx upper bound ([#7009](https://github.com/rapidsai/cuml/pull/7009)) [@bdice](https://github.com/bdice)
- Compatibility for dask ucx protocol changes ([#7008](https://github.com/rapidsai/cuml/pull/7008)) [@TomAugspurger](https://github.com/TomAugspurger)
- Add "devcontainers" dependency file key that excludes xgboost ([#7006](https://github.com/rapidsai/cuml/pull/7006)) [@bdice](https://github.com/bdice)
- Use `cuvs::neighbors::knn_merge_parts` ([#7005](https://github.com/rapidsai/cuml/pull/7005)) [@jcrist](https://github.com/jcrist)
- Mark `Ridge` as not supporting underdetermined systems ([#7003](https://github.com/rapidsai/cuml/pull/7003)) [@betatim](https://github.com/betatim)
- Reduce `arima` kernels binary size ([#6997](https://github.com/rapidsai/cuml/pull/6997)) [@jcrist](https://github.com/jcrist)
- Run the upstream HDBSCAN tests in CI ([#6995](https://github.com/rapidsai/cuml/pull/6995)) [@jcrist](https://github.com/jcrist)
- Use `pytest.ini` for all upstream tests ([#6994](https://github.com/rapidsai/cuml/pull/6994)) [@jcrist](https://github.com/jcrist)
- Support invoking `cuml_accel_tests` without extra pytest flags ([#6993](https://github.com/rapidsai/cuml/pull/6993)) [@jcrist](https://github.com/jcrist)
- Fall back on unsupported metrics in TSNE ([#6992](https://github.com/rapidsai/cuml/pull/6992)) [@jcrist](https://github.com/jcrist)
- Expand `LogisticRegression` limitations ([#6991](https://github.com/rapidsai/cuml/pull/6991)) [@jcrist](https://github.com/jcrist)
- Expand `sklearn.ensemble` limitations ([#6990](https://github.com/rapidsai/cuml/pull/6990)) [@jcrist](https://github.com/jcrist)
- Run `umap-learn` test suite in CI ([#6989](https://github.com/rapidsai/cuml/pull/6989)) [@jcrist](https://github.com/jcrist)
- Add a CI run with "intermediate" versions of dependencies ([#6988](https://github.com/rapidsai/cuml/pull/6988)) [@betatim](https://github.com/betatim)
- Correctly handle non-default GPU context in FIL ([#6987](https://github.com/rapidsai/cuml/pull/6987)) [@hcho3](https://github.com/hcho3)
- Move `cuml.accel` upstream tests to the `cuml_accel_tests` directory ([#6986](https://github.com/rapidsai/cuml/pull/6986)) [@jcrist](https://github.com/jcrist)
- Compute `components_` in `DBSCAN` ([#6976](https://github.com/rapidsai/cuml/pull/6976)) [@jcrist](https://github.com/jcrist)
- Add the cython-lint commit to `.git-blame-ignore-revs` ([#6972](https://github.com/rapidsai/cuml/pull/6972)) [@jcrist](https://github.com/jcrist)
- Several SVC cleanups ([#6970](https://github.com/rapidsai/cuml/pull/6970)) [@jcrist](https://github.com/jcrist)
- Update cython-lint, fix long lines ([#6969](https://github.com/rapidsai/cuml/pull/6969)) [@jcrist](https://github.com/jcrist)
- Enhance logging for cuml.accel proxy estimators ([#6957](https://github.com/rapidsai/cuml/pull/6957)) [@csadorf](https://github.com/csadorf)
- Add a reason to all `UnsupportedOnGPU`/`UnsupportedOnCPU` errors ([#6954](https://github.com/rapidsai/cuml/pull/6954)) [@jcrist](https://github.com/jcrist)
- Deprecating `data_on_host` parameter for UMAP ([#6953](https://github.com/rapidsai/cuml/pull/6953)) [@jinsolp](https://github.com/jinsolp)
- Implement metadata routing in `cuml.accel` ([#6950](https://github.com/rapidsai/cuml/pull/6950)) [@jcrist](https://github.com/jcrist)
- xfail some tests due to new scipy 1.16 release ([#6946](https://github.com/rapidsai/cuml/pull/6946)) [@jcrist](https://github.com/jcrist)
- Use CUDA 12.9 in Conda, Devcontainers, Spark, GHA, etc. ([#6944](https://github.com/rapidsai/cuml/pull/6944)) [@jakirkham](https://github.com/jakirkham)
- Deprecate hdbscan prediction functions in `cuml.cluster` namespace ([#6943](https://github.com/rapidsai/cuml/pull/6943)) [@jcrist](https://github.com/jcrist)
- Support `set_output` and `get_feature_names_out` in `cuml.accel` ([#6942](https://github.com/rapidsai/cuml/pull/6942)) [@jcrist](https://github.com/jcrist)
- Reflect output type in `hdbscan` prediction functions ([#6939](https://github.com/rapidsai/cuml/pull/6939)) [@jcrist](https://github.com/jcrist)
- Remove `Base.__getattr__`/`Base.__getstate__`/`Base.__setstate__` ([#6937](https://github.com/rapidsai/cuml/pull/6937)) [@jcrist](https://github.com/jcrist)
- Deprecate `connectivity` parameter to `HDBSCAN` ([#6936](https://github.com/rapidsai/cuml/pull/6936)) [@jcrist](https://github.com/jcrist)
- Remove nvidia and dask channels ([#6935](https://github.com/rapidsai/cuml/pull/6935)) [@vyasr](https://github.com/vyasr)
- Fixup KernelRidge hypothesis test ([#6931](https://github.com/rapidsai/cuml/pull/6931)) [@jcrist](https://github.com/jcrist)
- A few fixes for CI ([#6928](https://github.com/rapidsai/cuml/pull/6928)) [@jcrist](https://github.com/jcrist)
- Deprecate `penalty='none'` in `MBSGDClassifier`/`MBSGDRegressor`/`SGD` ([#6926](https://github.com/rapidsai/cuml/pull/6926)) [@jcrist](https://github.com/jcrist)
- Run FIL demo notebook in CI ([#6923](https://github.com/rapidsai/cuml/pull/6923)) [@hcho3](https://github.com/hcho3)
- Restore old behavior for `threshold` in FIL ([#6922](https://github.com/rapidsai/cuml/pull/6922)) [@hcho3](https://github.com/hcho3)
- De-Cython several modules ([#6920](https://github.com/rapidsai/cuml/pull/6920)) [@jcrist](https://github.com/jcrist)
- Remove `UniversalBase` ([#6919](https://github.com/rapidsai/cuml/pull/6919)) [@jcrist](https://github.com/jcrist)
- Add `KernelRidge` to `cuml.accel` ([#6917](https://github.com/rapidsai/cuml/pull/6917)) [@jcrist](https://github.com/jcrist)
- Port `HDBSCAN` to `InteropMixin`/`ProxyBase` ([#6916](https://github.com/rapidsai/cuml/pull/6916)) [@jcrist](https://github.com/jcrist)
- Expose `n_iter_` for `LogisticRegression` models ([#6911](https://github.com/rapidsai/cuml/pull/6911)) [@betatim](https://github.com/betatim)
- Warn and fail on stale xfail list entries ([#6907](https://github.com/rapidsai/cuml/pull/6907)) [@csadorf](https://github.com/csadorf)
- Handling all identical vectors in UMAP `smooth_knn_dist_kernel` ([#6904](https://github.com/rapidsai/cuml/pull/6904)) [@jinsolp](https://github.com/jinsolp)
- refactor(shellcheck): fix all shellcheck warnings/errors ([#6901](https://github.com/rapidsai/cuml/pull/6901)) [@gforsyth](https://github.com/gforsyth)
- Instantiate only specific RAFT linewise kernels ([#6900](https://github.com/rapidsai/cuml/pull/6900)) [@aamijar](https://github.com/aamijar)
- Remove `experimental.fil` Python module ([#6899](https://github.com/rapidsai/cuml/pull/6899)) [@hcho3](https://github.com/hcho3)
- Remove `GlobalSettings.device_type` ([#6895](https://github.com/rapidsai/cuml/pull/6895)) [@jcrist](https://github.com/jcrist)
- Add xfail list auto-formatter ([#6892](https://github.com/rapidsai/cuml/pull/6892)) [@csadorf](https://github.com/csadorf)
- Remove deprecated device selection functions ([#6888](https://github.com/rapidsai/cuml/pull/6888)) [@jcrist](https://github.com/jcrist)
- Remove deprecations for 25.08 ([#6887](https://github.com/rapidsai/cuml/pull/6887)) [@jcrist](https://github.com/jcrist)
- Removing tests, codepaths and other bits related to CUDA 11 ([#6885](https://github.com/rapidsai/cuml/pull/6885)) [@dantegd](https://github.com/dantegd)
- Change `KMeans` `random_state` default to `None` ([#6884](https://github.com/rapidsai/cuml/pull/6884)) [@jcrist](https://github.com/jcrist)
- Remove cuvs symbols in libcuml headers ([#6883](https://github.com/rapidsai/cuml/pull/6883)) [@hcho3](https://github.com/hcho3)
- Migrate C++ files from `src/experimental/fil` to `src/fil` ([#6882](https://github.com/rapidsai/cuml/pull/6882)) [@hcho3](https://github.com/hcho3)
- Remove invalid xfail entry and mark correct test as flaky ([#6877](https://github.com/rapidsai/cuml/pull/6877)) [@csadorf](https://github.com/csadorf)
- A few `max_features` fixups for `cuml.ensemble` ([#6873](https://github.com/rapidsai/cuml/pull/6873)) [@jcrist](https://github.com/jcrist)
- Loosen constrain on numpy upper version ([#6871](https://github.com/rapidsai/cuml/pull/6871)) [@csadorf](https://github.com/csadorf)
- Correct `sys.argv` when running `cuml.accel` ([#6870](https://github.com/rapidsai/cuml/pull/6870)) [@jcrist](https://github.com/jcrist)
- Support `LinearSVC` and `LinearSVR` in `cuml.accel` ([#6866](https://github.com/rapidsai/cuml/pull/6866)) [@viclafargue](https://github.com/viclafargue)
- Port `cuml.ensemble` to `InteropMixin`/`ProxyBase` ([#6863](https://github.com/rapidsai/cuml/pull/6863)) [@jcrist](https://github.com/jcrist)
- Deprecate some `cuml.accel` CLI options ([#6855](https://github.com/rapidsai/cuml/pull/6855)) [@jcrist](https://github.com/jcrist)
- Remove deprecated args in `cuml.ensemble` ([#6854](https://github.com/rapidsai/cuml/pull/6854)) [@jcrist](https://github.com/jcrist)
- Move deprecated positional/keyword args to keyword-only ([#6853](https://github.com/rapidsai/cuml/pull/6853)) [@jcrist](https://github.com/jcrist)
- Support `-c` and `-` options in `cuml.accel` CLI ([#6852](https://github.com/rapidsai/cuml/pull/6852)) [@jcrist](https://github.com/jcrist)
- Port `cuml.neighbors` to `InteropMixin`/`ProxyBase` ([#6851](https://github.com/rapidsai/cuml/pull/6851)) [@jcrist](https://github.com/jcrist)
- Remove CUDA 11 from dependencies.yaml ([#6847](https://github.com/rapidsai/cuml/pull/6847)) [@KyleFromNVIDIA](https://github.com/KyleFromNVIDIA)
- Add support for scikit-learn version 1.4 ([#6845](https://github.com/rapidsai/cuml/pull/6845)) [@betatim](https://github.com/betatim)
- Remove legacy FIL from libcuml ([#6844](https://github.com/rapidsai/cuml/pull/6844)) [@hcho3](https://github.com/hcho3)
- Port UMAP to InteropMixin/ProxyBase ([#6840](https://github.com/rapidsai/cuml/pull/6840)) [@jcrist](https://github.com/jcrist)
- Move FIL to subclass from Base ([#6835](https://github.com/rapidsai/cuml/pull/6835)) [@jcrist](https://github.com/jcrist)
- Mark `test_equivalence_components_pca_spca` as flaky ([#6834](https://github.com/rapidsai/cuml/pull/6834)) [@jcrist](https://github.com/jcrist)
- Fix several bugs in `cuml.accel` CLI ([#6832](https://github.com/rapidsai/cuml/pull/6832)) [@jcrist](https://github.com/jcrist)
- stop uploading packages to downloads.rapids.ai (pt 2) ([#6828](https://github.com/rapidsai/cuml/pull/6828)) [@jameslamb](https://github.com/jameslamb)
- Remove CUDA 11 devcontainers and update CI scripts ([#6818](https://github.com/rapidsai/cuml/pull/6818)) [@bdice](https://github.com/bdice)
- refactor(rattler): remove cuda11 options and general cleanup ([#6817](https://github.com/rapidsai/cuml/pull/6817)) [@gforsyth](https://github.com/gforsyth)
- Add GitHub Actions workflow to assign PRs to their authors ([#6816](https://github.com/rapidsai/cuml/pull/6816)) [@csadorf](https://github.com/csadorf)
- stop uploading packages to downloads.rapids.ai ([#6803](https://github.com/rapidsai/cuml/pull/6803)) [@jameslamb](https://github.com/jameslamb)
- Instantiate only specific RAFT reduction kernels ([#6780](https://github.com/rapidsai/cuml/pull/6780)) [@divyegala](https://github.com/divyegala)
- Forward-merge branch-25.06 into branch-25.08 ([#6647](https://github.com/rapidsai/cuml/pull/6647)) [@gforsyth](https://github.com/gforsyth)
- [FEA] HDBSCAN to Migrate to cuVS cluster prims from raft::cluster ([#6560](https://github.com/rapidsai/cuml/pull/6560)) [@tarang-jain](https://github.com/tarang-jain)

# cuml 25.06.00 (5 Jun 2025)

## 🚨 Breaking Changes
Expand Down
36 changes: 20 additions & 16 deletions python/cuml/tests/test_spectral_embedding.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,17 +16,12 @@
import cupy as cp
import numpy as np
import pytest
from sklearn.datasets import (
fetch_openml,
load_digits,
make_s_curve,
make_swiss_roll,
)
from sklearn.datasets import load_digits, make_s_curve, make_swiss_roll
from sklearn.manifold import SpectralEmbedding as skSpectralEmbedding
from sklearn.manifold import trustworthiness
from sklearn.model_selection import train_test_split

from cuml.manifold import SpectralEmbedding, spectral_embedding
from cuml.testing.datasets import make_classification_dataset

# Test parameters
N_NEIGHBORS = 15
Expand All @@ -45,15 +40,24 @@ def generate_swiss_roll(n_samples):
return X


def load_mnist(n_samples):
"""Load and sample MNIST dataset."""
mnist = fetch_openml("mnist_784", version=1, as_frame=False, parser="auto")
X, y = mnist.data, mnist.target.astype(np.int32)
X = X / 255.0
X, _, _, _ = train_test_split(
X, y, train_size=n_samples, stratify=y, random_state=42
def generate_mnist_like_dataset(n_samples):
"""Load and sample dataset using cuML's testing infrastructure."""

# Generate a classification dataset with similar characteristics to MNIST
# MNIST has 784 features (28x28 pixels) and 10 classes
X_train, X_test, y_train, y_test = make_classification_dataset(
datatype=np.float32,
nrows=n_samples,
ncols=784, # Same as MNIST features
n_info=100, # Number of informative features
num_classes=10, # Same as MNIST classes
)
return X

# Normalize to [0, 1] range like MNIST
X_train = (X_train - X_train.min()) / (X_train.max() - X_train.min())
X_test = (X_test - X_test.min()) / (X_test.max() - X_test.min())

return X_train


def load_digits_dataset(n_samples=None):
Expand All @@ -68,7 +72,7 @@ def load_digits_dataset(n_samples=None):
(generate_s_curve, 2000, 0.8),
(generate_swiss_roll, 2000, 0.8),
(generate_swiss_roll, 3000, 0.8),
(load_mnist, 5000, 0.8),
(generate_mnist_like_dataset, 5000, 0.8),
(load_digits_dataset, None, 0.8),
]

Expand Down