From 49fad58711ada22ef714ea31e08b824447d5da72 Mon Sep 17 00:00:00 2001
From: Samuel Moors <samuel.moors@vub.be>
Date: Fri, 20 Dec 2024 13:48:29 +0100
Subject: [PATCH] adding easyconfigs: ESM3-3.1.2-foss-2023a.eb,
 ESM3-3.1.2-foss-2023a-CUDA-12.1.1.eb

---
 .../ESM3/ESM3-3.1.2-foss-2023a-CUDA-12.1.1.eb | 68 +++++++++++++++++++
 .../e/ESM3/ESM3-3.1.2-foss-2023a.eb           | 68 +++++++++++++++++++
 2 files changed, 136 insertions(+)
 create mode 100644 easybuild/easyconfigs/e/ESM3/ESM3-3.1.2-foss-2023a-CUDA-12.1.1.eb
 create mode 100644 easybuild/easyconfigs/e/ESM3/ESM3-3.1.2-foss-2023a.eb

diff --git a/easybuild/easyconfigs/e/ESM3/ESM3-3.1.2-foss-2023a-CUDA-12.1.1.eb b/easybuild/easyconfigs/e/ESM3/ESM3-3.1.2-foss-2023a-CUDA-12.1.1.eb
new file mode 100644
index 00000000000..f3dbb7192f2
--- /dev/null
+++ b/easybuild/easyconfigs/e/ESM3/ESM3-3.1.2-foss-2023a-CUDA-12.1.1.eb
@@ -0,0 +1,68 @@
+easyblock = 'PythonBundle'
+
+name = 'ESM3'
+version = '3.1.2'
+versionsuffix = '-CUDA-%(cudaver)s'
+
+homepage = 'https://www.evolutionaryscale.ai/'
+description = """ESM3 is a frontier generative model for biology, able to jointly reason across
+three fundamental biological properties of proteins: sequence, structure, and
+function. These three data modalities are represented as tracks of discrete
+tokens at the input and output of ESM3. You can present the model with a
+combination of partial inputs across the tracks, and ESM3 will provide output
+predictions for all the tracks.
+"""
+
+toolchain = {'name': 'foss', 'version': '2023a'}
+
+builddependencies = [
+    ('hatch', '1.9.7'),  # for biotite
+]
+
+dependencies = [
+    ('CUDA', '12.1.1', '', SYSTEM),
+    ('Python', '3.11.3'),
+    ('PyTorch-bundle', '2.3.0', versionsuffix),
+    ('Transformers', '4.39.3'),
+    ('Biopython', '1.83'),
+    ('Brotli-python', '1.0.9'),
+    ('einops', '0.7.0'),
+    ('IPython', '8.14.0'),
+    ('scikit-learn', '1.3.1'),
+    ('Cython', '3.0.8'),  # for biotite
+    ('zstd', '1.5.5'),
+]
+
+use_pip = True
+
+exts_list = [
+    ('biotite', '0.41.2', {
+        'checksums': ['f4ebca8071fade51103c9e8534d60fff2021c22e6af85058a53ed6a3d96b6ad4'],
+    }),
+    ('msgpack-numpy', '0.4.8', {
+        'checksums': ['c667d3180513422f9c7545be5eec5d296dcbb357e06f72ed39cc683797556e69'],
+    }),
+    ('cloudpathlib', '0.16.0', {
+        'checksums': ['cdfcd35d46d529587d744154a0bdf962aca953b725c8784cd2ec478354ea63a3'],
+    }),
+    ('zstd', '1.5.5.1', {
+        'checksums': ['1ef980abf0e1e072b028d2d76ef95b476632651c96225cf30b619c6eef625672'],
+    }),
+    ('tenacity', '8.5.0', {
+        'checksums': ['8bc6c0c8a09b31e6cad13c47afbed1a567518250a9a171418582ed8d9c20ca78'],
+    }),
+    ('esm', version, {
+        'source_tmpl': '%(name)s-%(version)s-py3-none-any.whl',
+        'checksums': ['0c5ba0e12c28a1ddcad8a1e3fe152e476f2d24d412a8a04085b057f72759ba46'],
+    }),
+]
+
+sanity_pip_check = True
+
+sanity_check_commands = [
+    "python -c 'import esm'",
+    "python -c 'from esm.models.esm3 import ESM3'",
+    "python -c 'from esm.models.esmc import ESMC'",
+]
+
+moduleclass = 'bio'
diff --git a/easybuild/easyconfigs/e/ESM3/ESM3-3.1.2-foss-2023a.eb b/easybuild/easyconfigs/e/ESM3/ESM3-3.1.2-foss-2023a.eb
new file mode 100644
index 00000000000..39cf45e3530
--- /dev/null
+++ b/easybuild/easyconfigs/e/ESM3/ESM3-3.1.2-foss-2023a.eb
@@ -0,0 +1,68 @@
+easyblock = 'PythonBundle'
+
+name = 'ESM3'
+version = '3.1.2'
+
+homepage = 'https://www.evolutionaryscale.ai/'
+description = """ESM3 is a frontier generative model for biology, able to jointly reason across
+three fundamental biological properties of proteins: sequence, structure, and
+function. These three data modalities are represented as tracks of discrete
+tokens at the input and output of ESM3. You can present the model with a
+combination of partial inputs across the tracks, and ESM3 will provide output
+predictions for all the tracks.
+"""
+
+toolchain = {'name': 'foss', 'version': '2023a'}
+
+builddependencies = [
+    ('hatch', '1.9.7'),  # for biotite
+    # ('Ninja', '1.11.1'),
+]
+
+dependencies = [
+    ('Python', '3.11.3'),
+    ('PyTorch-bundle', '2.3.0'),
+    ('Transformers', '4.39.3'),
+    ('Biopython', '1.83'),
+    # ('Biotite', '0.41.0'),
+    ('Brotli-python', '1.0.9'),
+    ('einops', '0.7.0'),
+    ('IPython', '8.14.0'),
+    ('scikit-learn', '1.3.1'),
+    ('Cython', '3.0.8'),  # for biotite
+    ('zstd', '1.5.5'),
+]
+
+use_pip = True
+
+exts_list = [
+    ('biotite', '0.41.2', {
+        'checksums': ['f4ebca8071fade51103c9e8534d60fff2021c22e6af85058a53ed6a3d96b6ad4'],
+    }),
+    ('msgpack-numpy', '0.4.8', {
+        'checksums': ['c667d3180513422f9c7545be5eec5d296dcbb357e06f72ed39cc683797556e69'],
+    }),
+    ('cloudpathlib', '0.16.0', {
+        'checksums': ['cdfcd35d46d529587d744154a0bdf962aca953b725c8784cd2ec478354ea63a3'],
+    }),
+    ('zstd', '1.5.5.1', {
+        'checksums': ['1ef980abf0e1e072b028d2d76ef95b476632651c96225cf30b619c6eef625672'],
+    }),
+    ('tenacity', '8.5.0', {
+        'checksums': ['8bc6c0c8a09b31e6cad13c47afbed1a567518250a9a171418582ed8d9c20ca78'],
+    }),
+    ('esm', version, {
+        'source_tmpl': '%(name)s-%(version)s-py3-none-any.whl',
+        'checksums': ['0c5ba0e12c28a1ddcad8a1e3fe152e476f2d24d412a8a04085b057f72759ba46'],
+    }),
+]
+
+sanity_pip_check = True
+
+sanity_check_commands = [
+    "python -c 'import esm'",
+    "python -c 'from esm.models.esm3 import ESM3'",
+    "python -c 'from esm.models.esmc import ESMC'",
+]
+
+moduleclass = 'bio'