Merge pull request #1 from GT4SD/main

Merge

.clabot

@@ -1 +1 @@
-{ "contributors": [ "Ashish13898", "C-nit", "GitHub Action", "NicolaiRee", "O-Schilter", "YoelShoshan", "avaucher", "christofid", "donovat", "drugilsberg", "edux300", "elzinga-ibm-research", "federicozipoli", "georgosgeorgos", "helderlopes97", "helenaMontenegro", "hoffmansc", "jannisborn", "jrs021", "kishimoto1974", "kwehden", "mirunacrt", "polat-c", "yvesnana" ], "message": "Thanks a lot for working on GT4SD, we strongly value contributions from our users. It appears that one of the commiters in the PR did not sign the CLA for contributors. To do so, you can open an issue to sign the CLA clicking [here](https://github.com/GT4SD/gt4sd-core/issues/new?labels=cla-signing&template=cla-signature.yaml&title=CLA+signature). More details can be found [here](https://github.com/GT4SD/gt4sd-core/blob/main/CONTRIBUTING.md#contributor-license-agreement)." }
+{ "contributors": [ "Ashish13898", "C-nit", "GitHub Action", "NicolaiRee", "O-Schilter", "YoelShoshan", "avaucher", "christofid", "donovat", "drugilsberg", "edux300", "elzinga-ibm-research", "federicozipoli", "fiskrt", "georgosgeorgos", "helderlopes97", "helenaMontenegro", "hoffmansc", "jannisborn", "jrs021", "kishimoto1974", "kwehden", "mirunacrt", "polat-c", "yvesnana" ], "message": "Thanks a lot for working on GT4SD, we strongly value contributions from our users. It appears that one of the commiters in the PR did not sign the CLA for contributors. To do so, you can open an issue to sign the CLA clicking [here](https://github.com/GT4SD/gt4sd-core/issues/new?labels=cla-signing&template=cla-signature.yaml&title=CLA+signature). More details can be found [here](https://github.com/GT4SD/gt4sd-core/blob/main/CONTRIBUTING.md#contributor-license-agreement)." }

.github/workflows/cla-signature.yaml

@@ -34,12 +34,12 @@ jobs:
         with:
           branch: main
           directory: gt4sd-core-main
-          github_token: ${{ secrets.GH_TOKEN }}
+          github_token: ${{ secrets.GITHUB_TOKEN }}
       - name: Mark signature in the issue
         uses: andymckay/labeler@e6c4322d0397f3240f0e7e30a33b5c5df2d39e90
         with:
           add-labels: cla-signed
-          repo-token: ${{ secrets.GH_TOKEN }}
+          repo-token: ${{ secrets.GITHUB_TOKEN }}
       - name: Close issue (existing contributor)
         uses: peter-evans/close-issue@v1
         if: ${{ env.signing_commit_hash == '' }}

.github/workflows/docs.yaml

@@ -27,7 +27,7 @@ jobs:
           conda activate gt4sd
           pip install --no-deps .
           # make sure we can build docs using specific dependencies
-          pip install "attrs==20.3.0" "markdown-it-py==0.6.2"
+          pip install "attrs==20.3.0" "markdown-it-py==1.0.0"
           cd docs && make html && cd ..
       - name: Commit documentation changes
         run: |

.github/workflows/pypi.yaml

@@ -16,6 +16,7 @@ jobs:
       max-parallel: 3
       matrix:
         python-version:
+          - "3.10"
           - 3.8
           - 3.7
         os:

.github/workflows/tests.yaml

@@ -19,6 +19,8 @@ jobs:
     defaults:
       run:
         shell: bash -l {0} # for conda command
+    env:
+      GIT_CLONE_PROTECTION_ACTIVE: false
     steps:
       - uses: actions/checkout@v2
       - uses: conda-incubator/setup-miniconda@v2
@@ -48,7 +50,7 @@ jobs:
       - name: Check mypy
         run: |
           conda activate gt4sd
-          python -m mypy src/gt4sd
+          python -m mypy src/gt4sd --show-traceback
       - name: Run pytests
         run: |
           conda activate gt4sd

README.md

@@ -27,10 +27,10 @@ Almost all pretrained models are also available via `gradio`-powered [web apps](
 
 Currently `gt4sd` relies on:
 
-- python>=3.7,<3.9
-- pip>=19.1,<20.3
+- python>=3.7,<=3.10
+- pip==24.0
 
-We are actively working on relaxing these, so stay tuned or help us with this by [contributing](./CONTRIBUTING.md) to the project.
+If you need others, help us by [contributing](./CONTRIBUTING.md) to the project.
 
 ### Conda
 

conda_cpu_linux.yml

@@ -5,8 +5,9 @@ channels:
   - https://conda.anaconda.org/pyg
   - https://conda.anaconda.org/conda-forge
 dependencies:
-  - python>=3.7,<3.9
-  - pip>=19.1,<20.3
+  - setuptools==69.5.1
+  - python>=3.7,<3.11
+  - pip=24.0
   - pytorch>=1.0,<=1.12.1
   - cpuonly
   - pytorch-scatter<=2.0.9=*cu102*

conda_cpu_mac.yml

@@ -5,8 +5,9 @@ channels:
   - https://conda.anaconda.org/pyg
   - https://conda.anaconda.org/conda-forge
 dependencies:
-  - python>=3.7,<3.9
-  - pip>=19.1,<20.3
+  - setuptools==69.5.1
+  - python>=3.7,<3.11
+  - pip=24.0
   - pytorch>=1.0,<=1.12.1
   - cpuonly
   - pytorch-scatter<=2.0.9

conda_gpu.yml

@@ -5,8 +5,9 @@ channels:
   - https://conda.anaconda.org/pyg
   - https://conda.anaconda.org/conda-forge
 dependencies:
-  - python>=3.7,<3.9
-  - pip>=19.1,<20.3
+  - setuptools==69.5.1
+  - python>=3.7,<3.11
+  - pip=24.0
   - pytorch>=1.0,<=1.12.1=*cu*
   - pytorch-scatter<=2.0.9=*cu102*
   - torchvision<=0.13.1=*cu*

dev_requirements.txt

@@ -1,16 +1,15 @@
-better-apidoc==0.3.1
+better-apidoc==0.3.2
 black==22.3.0
 docutils==0.17.1
 flake8==3.8.4
 flask==1.1.2
 flask_login==0.5.0
 # isort==5.7.0
 licenseheaders==0.8.8
-mypy==0.950
-myst-parser==0.13.3
-pytest==6.1.1
+mypy>=1.0.0
+myst-parser==1.0.0
+pytest==6.2.5
 pytest-cov==2.10.1
-sphinx==3.4.3
+sphinx>=5
 sphinx-autodoc-typehints==1.11.1
 jinja2<3.1.0
-sphinx_rtd_theme==0.5.1

notebooks/discovery-demo.ipynb

@@ -16,6 +16,16 @@
    "id": "d6a65c61",
    "metadata": {},
    "outputs": [],
+   "source": [
+    "!pip install --upgrade protobuf"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "94cd4c8c",
+   "metadata": {},
+   "outputs": [],
    "source": [
     "from typing import List\n",
     "from rdkit import Chem\n",
@@ -2325,7 +2335,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "Python 3",
+   "display_name": "Python 3 (ipykernel)",
    "language": "python",
    "name": "python3"
   },

requirements.txt

@@ -1,4 +1,5 @@
 # pypi requirements
+setuptools==69.5.1
 accelerate>=0.12,<0.20.0
 datasets>=1.11.0
 diffusers<=0.6.0
@@ -11,26 +12,28 @@ gt4sd-molformer>=0.1.1
 gt4sd-trainer-hf-pl>=0.0.2
 keras>=2.3.1,<2.11.0
 keybert>=0.7.0
-markdown-it-py<3.0.0,>=2.2.0
+markdown-it-py
 minio==7.0.1
 modlamp>=4.0.0
 molgx>=0.22.0a1
 nglview>=3.0.3
 numpy>=1.16.5,<1.24.0
+pandas<=2.0.3
 protobuf<3.20
 pyarrow>=8.0.0
-pydantic>=1.7.3,<2.0.0
+pydantic>=2.0.0
 pymatgen>=2022.11.7
 PyTDC==0.3.7
 pytorch_lightning<=1.7.7
 pyyaml>=5.4.1
 rdkit>=2022.3.5
+rdkit-stubs>=0.7
 regex>=2.5.91
 reinvent-chemistry==0.0.38
 sacremoses>=0.0.41
 scikit-learn>=1.0.0,<1.3.0
 scikit-optimize>=0.8.1
-scipy>=1.0.0
+scipy>=1.0.0,<=1.11.0
 sentencepiece>=0.1.95
 sentence_transformers>1.0,<=2.2.2
 sympy>=1.10.1
@@ -46,3 +49,5 @@ transformers>=4.22.0,<=4.24.0
 typing_extensions>=3.7.4.3
 wheel>=0.26
 xgboost>=1.7.6
+sphinx_rtd_theme==0.5.1
+pydantic-settings>=2.0.0

setup.cfg

@@ -98,13 +98,20 @@ ignore = E203, E501, W503
 [mypy]
 check_untyped_defs = True
 plugins = pydantic.mypy
+exclude = ^src/gt4sd/frameworks/gflownet/envs
+
+[mypy-gt4sd.frameworks.gflownet.envs.*]
+follow_imports = skip
 
 [mypy-pytest.*]
 ignore_missing_imports = True
 
 [mypy-rdkit.*]
 ignore_missing_imports = True
 
+[mypy-rdkit-stubs.*]
+ignore_errors = True
+
 [mypy-setuptools.*]
 ignore_missing_imports = True
 
@@ -141,7 +148,6 @@ ignore_missing_imports = True
 [mypy-transformers.*]
 ignore_missing_imports = True
 
-# to avoid mypy from crashing (https://github.com/python/mypy/issues/11045)
 [mypy-transformers.trainer]
 check_untyped_defs = False
 
@@ -284,4 +290,7 @@ ignore_missing_imports = True
 ignore_missing_imports = True
 
 [mypy-xgboost.*]
+ignore_missing_imports = True
+
+[mypy-pydantic_settings.*]
 ignore_missing_imports = True

src/gt4sd/__init__.py

@@ -23,7 +23,7 @@
 #
 """Module initialization."""
 
-__version__ = "1.4.1"
+__version__ = "1.4.2"
 __name__ = "gt4sd"
 
 # NOTE: configure SSL to allow unverified contexts by default

src/gt4sd/algorithms/conditional_generation/regression_transformer/implementation.py

@@ -1013,7 +1013,7 @@ def validate_output(self, sequences: List[Any]) -> Tuple[List[Any], List[int]]:
             if isinstance(self.tokenizer.text_tokenizer, PolymerGraphTokenizer):
                 # Copolymer models require specific validation
                 return validate_molecules(
-                    pattern_list=list(zip(*sequences))[0],
+                    pattern_list=list(zip(*sequences))[0],  # type: ignore
                     input_type=MoleculeFormat.copolymer,
                 )
 
@@ -1086,9 +1086,11 @@ def filter_substructures(
                     f"{keep} is not a valid SMILES/SELFIES. Instead substructure filtering "
                     f"based on sequence alone can be done and is set to: {self.text_filtering}"
                 )
-            if keep not in self.substructures_to_keep and not Chem.MolFromSmiles(
+            if keep not in self.substructures_to_keep and not Chem.MolFromSmiles(  # type: ignore
                 self.target
-            ).HasSubstructMatch(subs_mol):
+            ).HasSubstructMatch(
+                subs_mol
+            ):
                 logger.info(
                     f"{keep} could not be identified in SMILES/SELFIES on text level AND no "
                     "substructure match occurred, hence it will be ignored"
@@ -1120,7 +1122,7 @@ def filter_substructures(
                         sane = False
                         break
                 else:
-                    if not mol.HasSubstructMatch(subs_mol):
+                    if not mol.HasSubstructMatch(subs_mol):  # type: ignore
                         # Desired substructure not found
                         sane = False
                         break

src/gt4sd/algorithms/conditional_generation/regression_transformer/utils.py

@@ -65,20 +65,20 @@ def filter_stubbed(
         Tuple of tuples of length 2 with filtered, generated molecule and its properties.
     """
 
-    seed = Chem.MolFromSmiles(target)
+    seed = Chem.MolFromSmiles(target)  # type: ignore
 
-    seed_atoms = len(list(seed.GetAtoms()))
-    seed_bonds = seed.GetNumBonds()
+    seed_atoms = len(list(seed.GetAtoms()))  # type: ignore
+    seed_bonds = seed.GetNumBonds()  # type: ignore
 
     smis: List[str] = []
     props: List[str] = []
     for smi, prop in property_sequences:
         if smi == "":
             continue
         try:
-            mol = Chem.MolFromSmiles(smi)
-            num_atoms = len(list(mol.GetAtoms()))
-            num_bonds = mol.GetNumBonds()
+            mol = Chem.MolFromSmiles(smi)  # type: ignore
+            num_atoms = len(list(mol.GetAtoms()))  # type: ignore
+            num_bonds = mol.GetNumBonds()  # type: ignore
 
             if num_atoms > (threshold * seed_atoms) and num_bonds > (
                 threshold * seed_bonds

src/gt4sd/algorithms/controlled_sampling/advanced_manufacturing/implementation/core.py

@@ -154,7 +154,7 @@ def __init__(
         targets: Dict[str, float],
         property_predictors: Dict[str, PropertyPredictor],
         representations: RepresentationsDict,
-        representation_order: List[str] = None,
+        representation_order: Optional[List[str]] = None,
         scalers: Optional[Dict[str, Scaler]] = None,
         weights: Optional[Dict[str, float]] = None,
         custom_score_function: Optional[

src/gt4sd/algorithms/core.py

@@ -25,7 +25,7 @@
 
 from __future__ import annotations
 
-import collections
+from collections.abc import Hashable
 import logging
 import os
 import shutil
@@ -233,7 +233,7 @@ def sample(self, number_of_items: int = 100) -> Iterator[S]:
                 try:
                     valid_item = self.configuration.validate_item(item)
                     # check if sample is hashable
-                    if not isinstance(item, collections.Hashable):
+                    if not isinstance(item, Hashable):
                         yield valid_item
                         item_set.add(str(index))
                     else:
@@ -623,9 +623,11 @@ def save_version_from_training_pipeline_arguments(
                 target_version,
             )
             filepaths_mapping = {
-                filename: source_filepath
-                if os.path.exists(source_filepath)
-                else os.path.join(source_missing_path, filename)
+                filename: (
+                    source_filepath
+                    if os.path.exists(source_filepath)
+                    else os.path.join(source_missing_path, filename)
+                )
                 for filename, source_filepath in filepaths_mapping.items()
             }
             logger.info(f"Saving artifacts into {target_path}...")
@@ -713,9 +715,11 @@ def upload_version_from_training_pipeline_arguments(
 
             # mapping between filenames and paths for a version.
             filepaths_mapping = {
-                filename: source_filepath
-                if os.path.exists(source_filepath)
-                else os.path.join(source_missing_path, filename)
+                filename: (
+                    source_filepath
+                    if os.path.exists(source_filepath)
+                    else os.path.join(source_missing_path, filename)
+                )
                 for filename, source_filepath in filepaths_mapping.items()
             }
 
@@ -908,8 +912,6 @@ def get_configuration_class_with_attributes(
 
 
 class PropertyPredictor(ABC, Generic[S, U]):
-    """TODO: Might be deprecated in future release."""
-
     def __init__(self, context: U) -> None:
         """Property predictor to investigate items.
 

src/gt4sd/algorithms/generation/diffusion/geodiff/core.py

@@ -253,8 +253,8 @@ def visualize_2d_input(self, data: Data) -> None:
         molSize = (450, 300)
         drawer = MD2.MolDraw2DSVG(molSize[0], molSize[1])
         drawer.DrawMolecule(mc)
-        drawer.FinishDrawing()
-        svg = drawer.GetDrawingText()
+        drawer.FinishDrawing()  # type: ignore
+        svg = drawer.GetDrawingText()  # type: ignore
         display(SVG(svg.replace("svg:", "")))
 
     def visualize_3d(self, mols_gen: List[Chem.Mol]) -> None: