{ai}[foss/2024a] AlphaFold3 v3.0.1-20250908, dm-haiku v0.0.14, Flax v0.10.2, ... w/ CUDA 12.6.0 by pavelToman · Pull Request #24177 · easybuilders/easybuild-easyconfigs

pavelToman · 2025-10-09T14:37:02Z

AlphaFold3 easyconfig and patches from: #22067
Requires:

This is another try to make AF3 works.
Added new jax-0.6.2-CUDA-12.6.0 as a dependency.
Also other dependencies with jax as dep updated:

Optax, Flax, dm-haiku, jaxtyping, jax-triton (with Triton 3.3.1)

resolves vscentrum/vsc-software-stack#477

…iku-0.0.14-foss-2024a-CUDA-12.6.0.eb, Flax-0.10.2-gfbf-2024a-CUDA-12.6.0.eb, Optax-0.2.4-gfbf-2024a-CUDA-12.6.0.eb, jax-triton-0.3.0-foss-2024a-CUDA-12.6.0.eb, jaxtyping-0.2.38-foss-2024a-CUDA-12.6.0.eb and patches: AlphaFold3-3.0.0_disable_fetch.patch, AlphaFold3-3.0.0.20250109_data_path.patch, AlphaFold3-3.0.0.20250109_relax_requirements.patch

pavelToman · 2025-10-09T14:44:31Z

For now AF3 could be installed without the run_alphafold_test.py - this needs models from google.
Tested only by run_alphafold_data_test.py and few smoketests

github-actions · 2025-10-09T14:59:03Z

Updated software `dm-haiku-0.0.14-foss-2024a-CUDA-12.6.0.eb`

Diff against dm-haiku-0.0.13-foss-2023a.eb

easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.13-foss-2023a.eb

diff --git a/easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.13-foss-2023a.eb b/easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.14-foss-2024a-CUDA-12.6.0.eb
index 9e73940419..4fa7a934c3 100644
--- a/easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.13-foss-2023a.eb
+++ b/easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.14-foss-2024a-CUDA-12.6.0.eb
@@ -1,29 +1,34 @@
+# update 0.0.12: Thomas Hoffmann (EMBL)
 easyblock = 'PythonBundle'
 
 name = 'dm-haiku'
-version = '0.0.13'
+version = '0.0.14'
+versionsuffix = '-CUDA-%(cudaver)s'
 
 homepage = 'https://github.com/deepmind/dm-haiku'
 description = """Haiku is a simple neural network library for JAX developed by some of the authors of Sonnet, a neural
 network library for TensorFlow."""
 
-toolchain = {'name': 'foss', 'version': '2023a'}
+toolchain = {'name': 'foss', 'version': '2024a'}
+
+builddependencies = [('poetry', '1.8.3')]
 
 dependencies = [
-    ('Python', '3.11.3'),
-    ('SciPy-bundle', '2023.07'),
-    ('jax', '0.4.25'),  # required by jmp, also provides absl-py
+    ('CUDA', '12.6.0', '', SYSTEM),
+    ('Python', '3.12.3'),
+    ('SciPy-bundle', '2024.05'),
+    ('jax', '0.6.2', versionsuffix),
+    ('Flax', '0.10.2', versionsuffix),
+    ('absl-py', '2.1.0'),
 ]
 
 exts_list = [
     ('jmp', '0.0.4', {
-        'source_tmpl': SOURCE_PY3_WHL,
-        'checksums': ['6aa7adbddf2bd574b28c7faf6e81a735eb11f53386447896909c6968dc36807d'],
+        'checksums': ['5dfeb0fd7c7a9f72a70fff0aab9d0cbfae32a809c02f4037ff3485ceb33e1730'],
     }),
     ('dm_haiku', version, {
         'modulename': 'haiku',
-        'source_tmpl': SOURCE_PY3_WHL,
-        'checksums': ['ee9562c68a059f146ad07f555ca591cb8c11ef751afecc38353863562bd23f43'],
+        'checksums': ['0c1d56ef20e1e04d8551f788ff8964167092c5d8051b7c17ed3f9ce60c0b0a59'],
     }),
 ]

Diff against dm-haiku-0.0.12-foss-2023a-CUDA-12.1.1.eb

easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.12-foss-2023a-CUDA-12.1.1.eb

diff --git a/easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.12-foss-2023a-CUDA-12.1.1.eb b/easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.14-foss-2024a-CUDA-12.6.0.eb
index 76672f6fb7..4fa7a934c3 100644
--- a/easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.12-foss-2023a-CUDA-12.1.1.eb
+++ b/easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.14-foss-2024a-CUDA-12.6.0.eb
@@ -2,44 +2,33 @@
 easyblock = 'PythonBundle'
 
 name = 'dm-haiku'
-version = '0.0.12'
+version = '0.0.14'
 versionsuffix = '-CUDA-%(cudaver)s'
 
 homepage = 'https://github.com/deepmind/dm-haiku'
 description = """Haiku is a simple neural network library for JAX developed by some of the authors of Sonnet, a neural
 network library for TensorFlow."""
 
-toolchain = {'name': 'foss', 'version': '2023a'}
+toolchain = {'name': 'foss', 'version': '2024a'}
+
+builddependencies = [('poetry', '1.8.3')]
 
 dependencies = [
-    ('Python', '3.11.3'),
-    ('SciPy-bundle', '2023.07'),
-    ('jax', '0.4.25', versionsuffix),  # required by jmp, also provides absl-py
-    ('PyYAML', '6.0'),
-    ('CUDA', '12.1.1', '', SYSTEM),
-    ('tensorstore', '0.1.65'),
-    ('protobuf-python', '4.24.0'),
-    ('Optax', '0.2.2', versionsuffix),
+    ('CUDA', '12.6.0', '', SYSTEM),
+    ('Python', '3.12.3'),
+    ('SciPy-bundle', '2024.05'),
+    ('jax', '0.6.2', versionsuffix),
+    ('Flax', '0.10.2', versionsuffix),
+    ('absl-py', '2.1.0'),
 ]
 
 exts_list = [
     ('jmp', '0.0.4', {
         'checksums': ['5dfeb0fd7c7a9f72a70fff0aab9d0cbfae32a809c02f4037ff3485ceb33e1730'],
     }),
-    ('flax', '0.8.4', {
-        'checksums': ['968683f850198e1aa5eb2d9d1e20bead880ef7423c14f042db9d60848cb1c90b'],
-    }),
-    ('nest_asyncio', '1.6.0', {
-        'checksums': ['6f172d5449aca15afd6c646851f4e31e02c598d553a667e38cafa997cfec55fe'],
-    }),
-    ('orbax_checkpoint', '0.5.18', {
-        'modulename': 'orbax.checkpoint',
-        'preinstallopts': """sed -i 's/jax >= 0.4.25/&\\*/g' pyproject.toml &&""",
-        'checksums': ['29f5d311b412760bd6a2fecab3bdbf75407bc00dc6d0457d19478258ecc8fa6d'],
-    }),
-    (name, version, {
+    ('dm_haiku', version, {
         'modulename': 'haiku',
-        'checksums': ['ba0b3acf71433156737fe342c486da11727e5e6c9e054245f4f9b8f0b53eb608'],
+        'checksums': ['0c1d56ef20e1e04d8551f788ff8964167092c5d8051b7c17ed3f9ce60c0b0a59'],
     }),
 ]

Diff against dm-haiku-0.0.9-foss-2022a-CUDA-11.7.0.eb

easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.9-foss-2022a-CUDA-11.7.0.eb

diff --git a/easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.9-foss-2022a-CUDA-11.7.0.eb b/easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.14-foss-2024a-CUDA-12.6.0.eb
index eb44afa10f..4fa7a934c3 100644
--- a/easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.9-foss-2022a-CUDA-11.7.0.eb
+++ b/easybuild/easyconfigs/d/dm-haiku/dm-haiku-0.0.14-foss-2024a-CUDA-12.6.0.eb
@@ -1,29 +1,34 @@
+# update 0.0.12: Thomas Hoffmann (EMBL)
 easyblock = 'PythonBundle'
 
 name = 'dm-haiku'
-version = '0.0.9'
+version = '0.0.14'
 versionsuffix = '-CUDA-%(cudaver)s'
 
 homepage = 'https://github.com/deepmind/dm-haiku'
 description = """Haiku is a simple neural network library for JAX developed by some of the authors of Sonnet, a neural
 network library for TensorFlow."""
 
-toolchain = {'name': 'foss', 'version': '2022a'}
+toolchain = {'name': 'foss', 'version': '2024a'}
+
+builddependencies = [('poetry', '1.8.3')]
 
 dependencies = [
-    ('CUDA', '11.7.0', '', SYSTEM),
-    ('Python', '3.10.4'),
-    ('SciPy-bundle', '2022.05'),
-    ('jax', '0.3.25', versionsuffix),  # required by jmp, also provides absl-py
+    ('CUDA', '12.6.0', '', SYSTEM),
+    ('Python', '3.12.3'),
+    ('SciPy-bundle', '2024.05'),
+    ('jax', '0.6.2', versionsuffix),
+    ('Flax', '0.10.2', versionsuffix),
+    ('absl-py', '2.1.0'),
 ]
 
 exts_list = [
     ('jmp', '0.0.4', {
         'checksums': ['5dfeb0fd7c7a9f72a70fff0aab9d0cbfae32a809c02f4037ff3485ceb33e1730'],
     }),
-    (name, version, {
+    ('dm_haiku', version, {
         'modulename': 'haiku',
-        'checksums': ['97752b32cdbe5a3e2d1c60ea884d33eb4b36e7d410000f0d61b571417c925435'],
+        'checksums': ['0c1d56ef20e1e04d8551f788ff8964167092c5d8051b7c17ed3f9ce60c0b0a59'],
     }),
 ]

Updated software `Flax-0.10.2-gfbf-2024a-CUDA-12.6.0.eb`

Diff against Flax-0.8.4-gfbf-2023a-CUDA-12.1.1.eb

easybuild/easyconfigs/f/Flax/Flax-0.8.4-gfbf-2023a-CUDA-12.1.1.eb

diff --git a/easybuild/easyconfigs/f/Flax/Flax-0.8.4-gfbf-2023a-CUDA-12.1.1.eb b/easybuild/easyconfigs/f/Flax/Flax-0.10.2-gfbf-2024a-CUDA-12.6.0.eb
index 802e02583f..06b1736b90 100644
--- a/easybuild/easyconfigs/f/Flax/Flax-0.8.4-gfbf-2023a-CUDA-12.1.1.eb
+++ b/easybuild/easyconfigs/f/Flax/Flax-0.10.2-gfbf-2024a-CUDA-12.6.0.eb
@@ -1,7 +1,7 @@
 easyblock = 'PythonBundle'
 
 name = 'Flax'
-version = '0.8.4'
+version = '0.10.2'
 versionsuffix = '-CUDA-%(cudaver)s'
 
 homepage = 'https://flax.readthedocs.io'
@@ -9,31 +9,33 @@ description = """Flax is a high-performance neural network library and ecosystem
 designed for flexibility: Try new forms of training by forking an example and
 by modifying the training loop, not by adding features to a framework."""
 
-toolchain = {'name': 'gfbf', 'version': '2023a'}
+toolchain = {'name': 'gfbf', 'version': '2024a'}
 
+builddependencies = [('hatchling', '1.24.2')]
 dependencies = [
-    ('CUDA', '12.1.1', '', SYSTEM),
-    ('Python', '3.11.3'),
-    ('SciPy-bundle', '2023.07'),
-    ('jax', '0.4.25', versionsuffix),
-    ('Optax', '0.2.2', versionsuffix),
-    ('protobuf-python', '4.24.0'),
-    ('PyYAML', '6.0'),
+    ('CUDA', '12.6.0', '', SYSTEM),
+    ('Python', '3.12.3'),
+    ('SciPy-bundle', '2024.05'),
+    ('jax', '0.6.2', versionsuffix),
+    ('Optax', '0.2.4', versionsuffix),
+    ('protobuf-python', '5.28.0'),
+    ('PyYAML', '6.0.2'),
+    ('tensorstore', '0.1.72'),
 ]
 
 exts_list = [
     ('nest_asyncio', '1.6.0', {
         'checksums': ['6f172d5449aca15afd6c646851f4e31e02c598d553a667e38cafa997cfec55fe'],
     }),
-    ('orbax_checkpoint', '0.5.15', {
-        'modulename': 'orbax.checkpoint',
-        'checksums': ['15195e8d1b381b56f23a62a25599a3644f5d08655fa64f60bb1b938b8ffe7ef3'],
+    ('humanize', '4.11.0', {
+        'checksums': ['e66f36020a2d5a974c504bd2555cf770621dbdbb6d82f94a6857c0b1ea2608be'],
     }),
-    ('tensorstore', '0.1.60', {
-        'checksums': ['88da8f1978982101b8dbb144fd29ee362e4e8c97fc595c4992d555f80ce62a79'],
+    ('orbax_checkpoint', version, {
+        'modulename': 'orbax.checkpoint',
+        'checksums': ['e575ebe1f94e5cb6353ab8c9df81de0ca7cddc118645c3bfc17b8344f19d42f1'],
     }),
-    ('flax', '0.8.4', {
-        'checksums': ['968683f850198e1aa5eb2d9d1e20bead880ef7423c14f042db9d60848cb1c90b'],
+    ('%(namelower)s', version, {
+        'checksums': ['6f831350026ad48182ba6588bb4dd72dc1084985d9aca923254cb3e4c78d75f3'],
     }),
 ]

Diff against Flax-0.8.4-gfbf-2023a.eb

easybuild/easyconfigs/f/Flax/Flax-0.8.4-gfbf-2023a.eb

diff --git a/easybuild/easyconfigs/f/Flax/Flax-0.8.4-gfbf-2023a.eb b/easybuild/easyconfigs/f/Flax/Flax-0.10.2-gfbf-2024a-CUDA-12.6.0.eb
index fe69e5c053..06b1736b90 100644
--- a/easybuild/easyconfigs/f/Flax/Flax-0.8.4-gfbf-2023a.eb
+++ b/easybuild/easyconfigs/f/Flax/Flax-0.10.2-gfbf-2024a-CUDA-12.6.0.eb
@@ -1,37 +1,41 @@
 easyblock = 'PythonBundle'
 
 name = 'Flax'
-version = '0.8.4'
+version = '0.10.2'
+versionsuffix = '-CUDA-%(cudaver)s'
 
 homepage = 'https://flax.readthedocs.io'
 description = """Flax is a high-performance neural network library and ecosystem for JAX that is
 designed for flexibility: Try new forms of training by forking an example and
 by modifying the training loop, not by adding features to a framework."""
 
-toolchain = {'name': 'gfbf', 'version': '2023a'}
+toolchain = {'name': 'gfbf', 'version': '2024a'}
 
+builddependencies = [('hatchling', '1.24.2')]
 dependencies = [
-    ('Python', '3.11.3'),
-    ('SciPy-bundle', '2023.07'),
-    ('jax', '0.4.25'),
-    ('Optax', '0.2.2'),
-    ('protobuf-python', '4.24.0'),
-    ('PyYAML', '6.0'),
+    ('CUDA', '12.6.0', '', SYSTEM),
+    ('Python', '3.12.3'),
+    ('SciPy-bundle', '2024.05'),
+    ('jax', '0.6.2', versionsuffix),
+    ('Optax', '0.2.4', versionsuffix),
+    ('protobuf-python', '5.28.0'),
+    ('PyYAML', '6.0.2'),
+    ('tensorstore', '0.1.72'),
 ]
 
 exts_list = [
     ('nest_asyncio', '1.6.0', {
         'checksums': ['6f172d5449aca15afd6c646851f4e31e02c598d553a667e38cafa997cfec55fe'],
     }),
-    ('orbax_checkpoint', '0.5.15', {
-        'modulename': 'orbax.checkpoint',
-        'checksums': ['15195e8d1b381b56f23a62a25599a3644f5d08655fa64f60bb1b938b8ffe7ef3'],
+    ('humanize', '4.11.0', {
+        'checksums': ['e66f36020a2d5a974c504bd2555cf770621dbdbb6d82f94a6857c0b1ea2608be'],
     }),
-    ('tensorstore', '0.1.60', {
-        'checksums': ['88da8f1978982101b8dbb144fd29ee362e4e8c97fc595c4992d555f80ce62a79'],
+    ('orbax_checkpoint', version, {
+        'modulename': 'orbax.checkpoint',
+        'checksums': ['e575ebe1f94e5cb6353ab8c9df81de0ca7cddc118645c3bfc17b8344f19d42f1'],
     }),
-    ('flax', '0.8.4', {
-        'checksums': ['968683f850198e1aa5eb2d9d1e20bead880ef7423c14f042db9d60848cb1c90b'],
+    ('%(namelower)s', version, {
+        'checksums': ['6f831350026ad48182ba6588bb4dd72dc1084985d9aca923254cb3e4c78d75f3'],
     }),
 ]

Updated software `Optax-0.2.4-gfbf-2024a-CUDA-12.6.0.eb`

Diff against Optax-0.2.2-gfbf-2023a-CUDA-12.1.1.eb

easybuild/easyconfigs/o/Optax/Optax-0.2.2-gfbf-2023a-CUDA-12.1.1.eb

diff --git a/easybuild/easyconfigs/o/Optax/Optax-0.2.2-gfbf-2023a-CUDA-12.1.1.eb b/easybuild/easyconfigs/o/Optax/Optax-0.2.4-gfbf-2024a-CUDA-12.6.0.eb
index 45b587bc0d..4ab06352be 100644
--- a/easybuild/easyconfigs/o/Optax/Optax-0.2.2-gfbf-2023a-CUDA-12.1.1.eb
+++ b/easybuild/easyconfigs/o/Optax/Optax-0.2.4-gfbf-2024a-CUDA-12.6.0.eb
@@ -1,30 +1,30 @@
 easyblock = 'PythonBundle'
 
 name = 'Optax'
-version = '0.2.2'
+version = '0.2.4'
 versionsuffix = '-CUDA-%(cudaver)s'
 
 homepage = 'https://github.com/deepmind/optax'
 description = """Optax is a gradient processing and optimization library for JAX."""
 
-toolchain = {'name': 'gfbf', 'version': '2023a'}
+toolchain = {'name': 'gfbf', 'version': '2024a'}
 
 dependencies = [
-    ('CUDA', '12.1.1', '', SYSTEM),
-    ('Python', '3.11.3'),
-    ('SciPy-bundle', '2023.07'),
-    ('jax', '0.4.25', versionsuffix),
+    ('CUDA', '12.6.0', '', SYSTEM),
+    ('Python', '3.12.3'),
+    ('SciPy-bundle', '2024.05'),
+    ('jax', '0.6.2', versionsuffix),
 ]
 
 exts_list = [
-    ('toolz', '0.12.1', {
-        'checksums': ['ecca342664893f177a13dac0e6b41cbd8ac25a358e5f215316d43e2100224f4d'],
+    ('chex', '0.1.87', {
+        'checksums': ['0096d89cc8d898bb521ef4bfbf5c24549022b0e5b301f529ab57238896fe6c5d'],
     }),
-    ('chex', '0.1.86', {
-        'checksums': ['e8b0f96330eba4144659e1617c0f7a57b161e8cbb021e55c6d5056c7378091d1'],
+    ('toolz', '1.0.0', {
+        'checksums': ['2c86e3d9a04798ac556793bced838816296a2f085017664e4995cb40a1047a02'],
     }),
     ('optax', version, {
-        'checksums': ['f09bf790ef4b09fb9c35f79a07594c6196a719919985f542dc84b0bf97812e0e'],
+        'checksums': ['4e05d3d5307e6dde4c319187ae36e6cd3a0c035d4ed25e9e992449a304f47336'],
     }),
 ]

Diff against Optax-0.2.2-gfbf-2023a.eb

easybuild/easyconfigs/o/Optax/Optax-0.2.2-gfbf-2023a.eb

diff --git a/easybuild/easyconfigs/o/Optax/Optax-0.2.2-gfbf-2023a.eb b/easybuild/easyconfigs/o/Optax/Optax-0.2.4-gfbf-2024a-CUDA-12.6.0.eb
index 007b56ab55..4ab06352be 100644
--- a/easybuild/easyconfigs/o/Optax/Optax-0.2.2-gfbf-2023a.eb
+++ b/easybuild/easyconfigs/o/Optax/Optax-0.2.4-gfbf-2024a-CUDA-12.6.0.eb
@@ -1,28 +1,30 @@
 easyblock = 'PythonBundle'
 
 name = 'Optax'
-version = '0.2.2'
+version = '0.2.4'
+versionsuffix = '-CUDA-%(cudaver)s'
 
 homepage = 'https://github.com/deepmind/optax'
 description = """Optax is a gradient processing and optimization library for JAX."""
 
-toolchain = {'name': 'gfbf', 'version': '2023a'}
+toolchain = {'name': 'gfbf', 'version': '2024a'}
 
 dependencies = [
-    ('Python', '3.11.3'),
-    ('SciPy-bundle', '2023.07'),
-    ('jax', '0.4.25'),
+    ('CUDA', '12.6.0', '', SYSTEM),
+    ('Python', '3.12.3'),
+    ('SciPy-bundle', '2024.05'),
+    ('jax', '0.6.2', versionsuffix),
 ]
 
 exts_list = [
-    ('toolz', '0.12.1', {
-        'checksums': ['ecca342664893f177a13dac0e6b41cbd8ac25a358e5f215316d43e2100224f4d'],
+    ('chex', '0.1.87', {
+        'checksums': ['0096d89cc8d898bb521ef4bfbf5c24549022b0e5b301f529ab57238896fe6c5d'],
     }),
-    ('chex', '0.1.86', {
-        'checksums': ['e8b0f96330eba4144659e1617c0f7a57b161e8cbb021e55c6d5056c7378091d1'],
+    ('toolz', '1.0.0', {
+        'checksums': ['2c86e3d9a04798ac556793bced838816296a2f085017664e4995cb40a1047a02'],
     }),
     ('optax', version, {
-        'checksums': ['f09bf790ef4b09fb9c35f79a07594c6196a719919985f542dc84b0bf97812e0e'],
+        'checksums': ['4e05d3d5307e6dde4c319187ae36e6cd3a0c035d4ed25e9e992449a304f47336'],
     }),
 ]

Diff against Optax-0.1.7-foss-2022a-CUDA-11.7.0.eb

easybuild/easyconfigs/o/Optax/Optax-0.1.7-foss-2022a-CUDA-11.7.0.eb

diff --git a/easybuild/easyconfigs/o/Optax/Optax-0.1.7-foss-2022a-CUDA-11.7.0.eb b/easybuild/easyconfigs/o/Optax/Optax-0.2.4-gfbf-2024a-CUDA-12.6.0.eb
index 970a86cb00..4ab06352be 100644
--- a/easybuild/easyconfigs/o/Optax/Optax-0.1.7-foss-2022a-CUDA-11.7.0.eb
+++ b/easybuild/easyconfigs/o/Optax/Optax-0.2.4-gfbf-2024a-CUDA-12.6.0.eb
@@ -1,40 +1,33 @@
 easyblock = 'PythonBundle'
 
 name = 'Optax'
-version = '0.1.7'
+version = '0.2.4'
 versionsuffix = '-CUDA-%(cudaver)s'
 
 homepage = 'https://github.com/deepmind/optax'
 description = """Optax is a gradient processing and optimization library for JAX."""
 
-toolchain = {'name': 'foss', 'version': '2022a'}
+toolchain = {'name': 'gfbf', 'version': '2024a'}
 
 dependencies = [
-    ('CUDA', '11.7.0', '', SYSTEM),
-    ('Python', '3.10.4'),
-    ('dm-haiku', '0.0.9', versionsuffix),
-    ('dm-tree', '0.1.8'),
-    ('typing-extensions', '4.3.0'),
-]
-
-builddependencies = [
-    ('pytest-xdist', '2.5.0'),
+    ('CUDA', '12.6.0', '', SYSTEM),
+    ('Python', '3.12.3'),
+    ('SciPy-bundle', '2024.05'),
+    ('jax', '0.6.2', versionsuffix),
 ]
 
 exts_list = [
-    ('toolz', '0.12.0', {
-        'checksums': ['88c570861c440ee3f2f6037c4654613228ff40c93a6c25e0eba70d17282c6194'],
+    ('chex', '0.1.87', {
+        'checksums': ['0096d89cc8d898bb521ef4bfbf5c24549022b0e5b301f529ab57238896fe6c5d'],
     }),
-    ('chex', '0.1.6', {
-        'checksums': ['adb5d2352b5f0d248ccf594be1b1bf9ee7a2bee2a57f0eac78547538d479b0e7'],
+    ('toolz', '1.0.0', {
+        'checksums': ['2c86e3d9a04798ac556793bced838816296a2f085017664e4995cb40a1047a02'],
     }),
     ('optax', version, {
-        'checksums': ['6a5a848bc5e55e619b187c749fdddc4a5443ea14be85cc769f995779865c110d'],
-        # ignore equivalence_test.py which imports flax, which depends on optax
-        'runtest': 'python -s -m pytest -n %(parallel)s --pyargs --ignore-glob="**/equivalence_test.py" optax',
+        'checksums': ['4e05d3d5307e6dde4c319187ae36e6cd3a0c035d4ed25e9e992449a304f47336'],
     }),
 ]
 
-sanity_check_commands = ["python -s -c 'from optax import GradientTransformation'"]
+sanity_check_commands = ["python -c 'from optax import GradientTransformation'"]
 
 moduleclass = 'lib'

pavelToman · 2025-10-09T15:26:29Z

Test report by @pavelToman
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
node4006.donphan.os - Linux RHEL 9.6, x86_64, Intel(R) Xeon(R) Gold 6240 CPU @ 2.60GHz, 1 x NVIDIA NVIDIA A2, 580.82.07, Python 3.9.21
See https://gist.github.com/pavelToman/6a41cec5261cb12c00d5b2aca148954f for a full test report.

pavelToman · 2025-10-13T08:15:44Z

For now it seems that Triton-3-3.1 and jax-triton-0.3.0 are too recent for AF3.0.1.
Error - the kernel masks with 0xFFFFFFFF after an int32 bitcast; Triton ≥3.3 now rejects that constant as signed int32.

ValueError('Scalar 4294967295 is out of range for type int32')
... qk = qk.to(tl.int32, bitcast=True) & 0xFFFFFFFF

I will try to patch this or will go for jax-triton-0.2.0 with Triton-3.1.0.

ThomasHoffmann77 · 2025-10-14T14:28:57Z

@pavelToman there are some more recent untagged AF3 commits. We recently had the request to install google-deepmind/alphafold3@751a4b8 in order to reduce reduce peak memory usage of the jackhmmer. We are currently running commit a8ecdb2 together with a patched HMMER module (docker/jackhmmer_seq_limit.patch‎).

pavelToman · 2025-10-14T14:41:55Z

@ThomasHoffmann77 thanks for sharing this, I will try the latest commit a8ecdb2 then.
What about the run_alphafold_test.py? Does this one pass ok for you?

ThomasHoffmann77 · 2025-10-14T15:40:25Z

@ThomasHoffmann77 thanks for sharing this, I will try the latest commit a8ecdb2 then. What about the run_alphafold_test.py? Does this one pass ok for you?

I'll do a testrun and let you know.
btw, there are some AF3 commits about Triton incompatibility and one fixing an issue in run_alphafold_test.py

pavelToman · 2025-10-15T14:06:28Z

The latest commit a8ecdb2 wroks for me and run_alphafold_test.py pass!

pavelToman · 2025-10-15T15:07:36Z

Test report by @pavelToman
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
node4002.donphan.os - Linux RHEL 9.6, x86_64, Intel(R) Xeon(R) Gold 6240 CPU @ 2.60GHz, 1 x NVIDIA NVIDIA A2, 580.82.07, Python 3.9.21
See https://gist.github.com/pavelToman/8e87aa7bb26fc38693c33cd3b76d3ce0 for a full test report.

pavelToman · 2025-10-15T15:21:54Z

Test report by @pavelToman
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
node3906.accelgor.os - Linux RHEL 9.6, x86_64, AMD EPYC 7413 24-Core Processor, 1 x NVIDIA NVIDIA A100-SXM4-80GB, 580.82.07, Python 3.9.21
See https://gist.github.com/pavelToman/03ead2750b880c414530691832676bc8 for a full test report.

pavelToman · 2025-10-15T15:23:34Z

Test report by @pavelToman
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
node4304.litleo.os - Linux RHEL 9.6, x86_64, AMD EPYC 9454P 48-Core Processor, 1 x NVIDIA NVIDIA H100 NVL, 580.82.07, Python 3.9.21
See https://gist.github.com/pavelToman/46bc6134ad7cf4adbcfd79a3e1f75dda for a full test report.

pavelToman · 2025-10-16T08:02:40Z

Failing check:

 ERROR: test_pr_python_packages (test.easyconfigs.easyconfigs.EasyConfigTest)
Several checks for easyconfigs that install (bundles of) Python packages.
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/opt/hostedtoolcache/Python/3.7.17/x64/lib/python3.7/site-packages/easybuild/framework/easyconfig/easyconfig.py", line 2176, in resolve_template
    value = value % tmpl_dict
KeyError: 'start_dir'

@Flamefire respond on Slack:
The template is missing in the test as it is a runtime value.
#24260 fixes the test

Now this is not failing in checks anymore

…old3301

ThomasHoffmann77 · 2025-10-20T07:24:45Z

The latest commit a8ecdb2 wroks for me and run_alphafold_test.py pass!

for the seq_limit functionality (google-deepmind/alphafold3@751a4b8) HMMER requires a patch. See PR #24301.

ThomasHoffmann77

('HMMER', '3.4', '-seqlimit'),

pavelToman · 2025-10-20T09:53:27Z

Test report by @pavelToman
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
node4301.litleo.os - Linux RHEL 9.6, x86_64, AMD EPYC 9454P 48-Core Processor, 1 x NVIDIA NVIDIA H100 NVL, 580.95.05, Python 3.9.21
See https://gist.github.com/pavelToman/d35408ab1b0f89c19b57b0f38062a10e for a full test report.

all tests from run_alphafold_test.py passed

pavelToman · 2025-10-20T13:07:14Z

@boegelbot please test @ jsc-zen3-a100

boegelbot · 2025-10-20T13:10:10Z

@pavelToman: Request for testing this PR well received on jsczen3l1.int.jsc-zen3.fz-juelich.de

PR test command 'if [[ develop != 'develop' ]]; then EB_BRANCH=develop ./easybuild_develop.sh 2> /dev/null 1>&2; EB_PREFIX=/home/boegelbot/easybuild/develop source init_env_easybuild_develop.sh; fi; EB_PR=24177 EB_ARGS= EB_CONTAINER= EB_REPO=easybuild-easyconfigs EB_BRANCH=develop /opt/software/slurm/bin/sbatch --job-name test_PR_24177 --ntasks=8 --partition=jsczen3g --gres=gpu:1 ~/boegelbot/eb_from_pr_upload_jsc-zen3.sh' executed!

exit code: 0
output:

Submitted batch job 8493

Test results coming soon (I hope)...

Details

- notification for comment with ID 3421993442 processed

Message to humans: this is just bookkeeping information for me,
it is of no use to you (unless you think I have a bug, which I don't).

boegelbot · 2025-10-20T13:18:06Z

Test report by @boegelbot
FAILED
Build succeeded for 3 out of 7 (6 easyconfigs in total)
jsczen3g1.int.jsc-zen3.fz-juelich.de - Linux Rocky Linux 9.6, x86_64, AMD EPYC-Milan Processor (zen3), 1 x NVIDIA NVIDIA A100 80GB PCIe, 580.95.05, Python 3.9.21
See https://gist.github.com/boegelbot/769f1d819d822da4891ced1ac47e1d1f for a full test report.

pavelToman · 2025-10-20T13:28:57Z

Test report by @boegelbot FAILED Build succeeded for 3 out of 7 (6 easyconfigs in total) jsczen3g1.int.jsc-zen3.fz-juelich.de - Linux Rocky Linux 9.6, x86_64, AMD EPYC-Milan Processor (zen3), 1 x NVIDIA NVIDIA A100 80GB PCIe, 580.95.05, Python 3.9.21 See https://gist.github.com/boegelbot/769f1d819d822da4891ced1ac47e1d1f for a full test report.

Failed on installation of tensorstore-0.1.72-gfbf-2024a.eb

It seems this needs to be merged first?

fix failing build for tensorstore 0.1.72 when using RPATH by passing $TMPDIR from host into Bazel sandbox #23139

pavelToman · 2025-12-12T16:37:21Z

@boegelbot please test @ jsc-zen3-a100
CORE_CNT=16

boegelbot · 2025-12-12T16:40:09Z

@pavelToman: Request for testing this PR well received on jsczen3l1.int.jsc-zen3.fz-juelich.de

PR test command 'if [[ develop != 'develop' ]]; then EB_BRANCH=develop ./easybuild_develop.sh 2> /dev/null 1>&2; EB_PREFIX=/home/boegelbot/easybuild/develop source init_env_easybuild_develop.sh; fi; EB_PR=24177 EB_ARGS= EB_CONTAINER= EB_REPO=easybuild-easyconfigs EB_BRANCH=develop /opt/software/slurm/bin/sbatch --job-name test_PR_24177 --ntasks="16" --partition=jsczen3g --gres=gpu:1 ~/boegelbot/eb_from_pr_upload_jsc-zen3.sh' executed!

exit code: 0
output:

Submitted batch job 9114

Test results coming soon (I hope)...

Details

- notification for comment with ID 3647306088 processed

Message to humans: this is just bookkeeping information for me,
it is of no use to you (unless you think I have a bug, which I don't).

boegelbot · 2025-12-12T16:47:53Z

Test report by @boegelbot
FAILED
Build succeeded for 3 out of 7 (total: 7 mins 4 secs) (6 easyconfigs in total)
jsczen3g1.int.jsc-zen3.fz-juelich.de - Linux Rocky Linux 9.6, x86_64, AMD EPYC-Milan Processor (zen3), 1 x NVIDIA NVIDIA A100 80GB PCIe, 580.95.05, Python 3.9.21
See https://gist.github.com/boegelbot/97da21c07f0f867db3c3091a08b327e4 for a full test report.

…asyconfigs into 20251009163650_new_pr_AlphaFold3301

easybuild/easyconfigs/a/AlphaFold3/AlphaFold3-3.0.1-20250908-foss-2024a-CUDA-12.6.0.eb

…oss-2024a-CUDA-12.6.0.eb Co-authored-by: Lara Ramona Peeters <49882639+laraPPr@users.noreply.github.com>

pavelToman added new update labels Oct 9, 2025

pavelToman marked this pull request as draft October 9, 2025 14:38

pavelToman changed the title ~~{ai}[foss/2024a] AlphaFold3 v3.0.1, dm-haiku v0.0.14, Flax v0.10.2, ... w/ CUDA 12.6.0~~ {ai}[foss/2024a] AlphaFold3 v3.0.1, dm-haiku v0.0.14, Flax v0.10.2, ... w/ CUDA 12.6.0 WIP Oct 9, 2025

Thyre added the 2024a issues & PRs related to 2024a common toolchains label Oct 9, 2025

pavelToman mentioned this pull request Oct 9, 2025

{ai,bio}[foss/2024a] AlphaFold3 v3.0.1, jax-triton v0.2.0, jaxtyping v0.2.38, dm-tree v0.1.8 w/ CUDA 12.6.0 DRAFT! #22067

Draft

15 tasks

pavelToman added new update 2024a issues & PRs related to 2024a common toolchains labels Oct 9, 2025

pavelToman mentioned this pull request Oct 10, 2025

AlphaFold 3 vscentrum/vsc-software-stack#477

Open

2 tasks

update to the latest commit version a8ecdb2

db6b330

pavelToman marked this pull request as ready for review October 15, 2025 14:03

pavelToman changed the title ~~{ai}[foss/2024a] AlphaFold3 v3.0.1, dm-haiku v0.0.14, Flax v0.10.2, ... w/ CUDA 12.6.0 WIP~~ {ai}[foss/2024a] AlphaFold3 v3.0.1-20250908, dm-haiku v0.0.14, Flax v0.10.2, ... w/ CUDA 12.6.0 Oct 15, 2025

Update AlphaFold3-3.0.1-20250908-foss-2024a-CUDA-12.6.0.eb - typo

e32e48c

pavelToman mentioned this pull request Oct 15, 2025

{bio}[foss/2024a] AlphaFold3 WIP #21864

Open

Merge branch 'easybuilders:develop' into 20251009163650_new_pr_AlphaF…

e963f87

…old3301

ThomasHoffmann77 suggested changes Oct 20, 2025

View reviewed changes

ThomasHoffmann77 mentioned this pull request Oct 20, 2025

add patch for HMMER v3.4 to add support for --seq_limit option #24301

Merged

Update AlphaFold3-3.0.1-20250908-foss-2024a-CUDA-12.6.0.eb

f4a11a1

Update AlphaFold3-3.0.1-20250908-foss-2024a-CUDA-12.6.0.eb

ce6da6e

Merge branch 'develop' of https://github.com/easybuilders/easybuild-e…

e10e24b

…asyconfigs into 20251009163650_new_pr_AlphaFold3301

laraPPr requested changes Feb 2, 2026

View reviewed changes

easybuild/easyconfigs/a/AlphaFold3/AlphaFold3-3.0.1-20250908-foss-2024a-CUDA-12.6.0.eb Outdated Show resolved Hide resolved

Update easybuild/easyconfigs/a/AlphaFold3/AlphaFold3-3.0.1-20250908-f…

1961e93

…oss-2024a-CUDA-12.6.0.eb Co-authored-by: Lara Ramona Peeters <49882639+laraPPr@users.noreply.github.com>

Conversation

pavelToman commented Oct 9, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pavelToman commented Oct 9, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions bot commented Oct 9, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Updated software dm-haiku-0.0.14-foss-2024a-CUDA-12.6.0.eb

Updated software Flax-0.10.2-gfbf-2024a-CUDA-12.6.0.eb

Updated software Optax-0.2.4-gfbf-2024a-CUDA-12.6.0.eb

Uh oh!

pavelToman commented Oct 9, 2025

Uh oh!

pavelToman commented Oct 13, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ThomasHoffmann77 commented Oct 14, 2025

Uh oh!

pavelToman commented Oct 14, 2025

Uh oh!

ThomasHoffmann77 commented Oct 14, 2025

Uh oh!

pavelToman commented Oct 15, 2025

Uh oh!

pavelToman commented Oct 15, 2025

Uh oh!

pavelToman commented Oct 15, 2025

Uh oh!

pavelToman commented Oct 15, 2025

Uh oh!

pavelToman commented Oct 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ThomasHoffmann77 commented Oct 20, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ThomasHoffmann77 left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

pavelToman commented Oct 20, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pavelToman commented Oct 20, 2025

Uh oh!

boegelbot commented Oct 20, 2025

Uh oh!

boegelbot commented Oct 20, 2025

Uh oh!

pavelToman commented Oct 20, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pavelToman commented Dec 12, 2025

Uh oh!

boegelbot commented Dec 12, 2025

Uh oh!

boegelbot commented Dec 12, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

pavelToman commented Oct 9, 2025 •

edited

Loading

pavelToman commented Oct 9, 2025 •

edited

Loading

github-actions bot commented Oct 9, 2025 •

edited

Loading

Updated software `dm-haiku-0.0.14-foss-2024a-CUDA-12.6.0.eb`

Updated software `Flax-0.10.2-gfbf-2024a-CUDA-12.6.0.eb`

Updated software `Optax-0.2.4-gfbf-2024a-CUDA-12.6.0.eb`

pavelToman commented Oct 13, 2025 •

edited

Loading

pavelToman commented Oct 16, 2025 •

edited

Loading

ThomasHoffmann77 commented Oct 20, 2025 •

edited

Loading

ThomasHoffmann77 left a comment •

edited

Loading

pavelToman commented Oct 20, 2025 •

edited

Loading

pavelToman commented Oct 20, 2025 •

edited

Loading