{geo}[foss/2025a] LightlySSL v1.5.22 w/ CUDA 12.8.0

easybuild/easyconfigs/l/LightlySSL/LightlySSL-1.5.22-foss-2025a-CUDA-12.8.0.eb

@@ -0,0 +1,43 @@
+easyblock = 'PythonBundle'
+
+name = 'LightlySSL'
+version = '1.5.22'
+versionsuffix = '-CUDA-%(cudaver)s'
+
+homepage = 'https://docs.lightly.ai/self-supervised-learning/'
+description = """
+LightlySSL is a computer vision framework for self-supervised learning.
+With LightlySSL you can train deep learning models using self-supervision. This means, that you
+don’t require any labels to train a model. LightlySSL has been built to help you understand and work
+with large unlabeled datasets. It is built on top of PyTorch and therefore fully compatible with
+other frameworks such as Fast.ai.
+"""
+
+toolchain = {'name': 'foss', 'version': '2025a'}
+
+dependencies = [
+    ('Python', '3.13.1'),
+    ('CUDA', '12.8.0', '', SYSTEM),
+    ('SciPy-bundle', '2025.06'),
+    ('PyTorch', '2.9.1', f'{versionsuffix}-whl'),
+    ('PyTorch-bundle', '2.9.1', f'{versionsuffix}-whl'),
+    ('PyTorch-Lightning', '2.6.1', versionsuffix),
+    # optional
+    ('pydantic', '2.11.7'),
+    ('matplotlib', '3.10.3'),
+    ('timm', '1.0.24', versionsuffix),
+]
+
+exts_list = [
+    ('aenum', '3.1.16', {
+        'checksums': ['bfaf9589bdb418ee3a986d85750c7318d9d2839c1b1a1d6fe8fc53ec201cf140'],
+    }),
+    ('lightly_utils', '0.0.2', {
+        'checksums': ['a351f3d600f0ab08d12f294725c6457ae000645cb0a1083d0845cb196ccfe698'],
+    }),
+    ('lightly', version, {
+        'checksums': ['7bb8939976ea5e8fabb20d049812700d51e44f6461f15510a784afd1252a3dbd'],
+    }),
+]
+
+moduleclass = 'geo'

easybuild/easyconfigs/l/LightlySSL/LightlySSL-1.5.22-foss-2025a.eb

@@ -0,0 +1,41 @@
+easyblock = 'PythonBundle'
+
+name = 'LightlySSL'
+version = '1.5.22'
+
+homepage = 'https://docs.lightly.ai/self-supervised-learning/'
+description = """
+LightlySSL is a computer vision framework for self-supervised learning.
+With LightlySSL you can train deep learning models using self-supervision. This means, that you
+don’t require any labels to train a model. LightlySSL has been built to help you understand and work
+with large unlabeled datasets. It is built on top of PyTorch and therefore fully compatible with
+other frameworks such as Fast.ai.
+"""
+
+toolchain = {'name': 'foss', 'version': '2025a'}
+
+dependencies = [
+    ('Python', '3.13.1'),
+    ('SciPy-bundle', '2025.06'),
+    ('PyTorch', '2.9.1', '-whl'),
+    ('PyTorch-bundle', '2.9.1', '-whl'),
+    ('PyTorch-Lightning', '2.6.1'),
+    # optional
+    ('pydantic', '2.11.7'),
+    ('matplotlib', '3.10.3'),
+    ('timm', '1.0.24'),
+]
+
+exts_list = [
+    ('aenum', '3.1.16', {
+        'checksums': ['bfaf9589bdb418ee3a986d85750c7318d9d2839c1b1a1d6fe8fc53ec201cf140'],
+    }),
+    ('lightly_utils', '0.0.2', {
+        'checksums': ['a351f3d600f0ab08d12f294725c6457ae000645cb0a1083d0845cb196ccfe698'],
+    }),
+    ('lightly', version, {
+        'checksums': ['7bb8939976ea5e8fabb20d049812700d51e44f6461f15510a784afd1252a3dbd'],
+    }),
+]
+
+moduleclass = 'geo'