Require srsly >=2.5.3 for Python 3.14 cloudpickle fix

Use git config insteadOf to authenticate private repo access via env var
rather than interpolating the token directly into the pip install URL,
which would expose it in process listings and potentially in logs.

.github/workflows/cibuildwheel.yml

@@ -7,9 +7,12 @@ on:
       # ** matches 'zero or more of any character'
       - 'release-v[0-9]+.[0-9]+.[0-9]+**'
       - 'prerelease-v[0-9]+.[0-9]+.[0-9]+**'
+
+permissions: {}
+
 jobs:
   build_wheels:
-    uses: explosion/gha-cibuildwheel/.github/workflows/cibuildwheel.yml@main
+    uses: explosion/gha-cibuildwheel/.github/workflows/cibuildwheel.yml@2c98f757f13d112cf73fcf4b627249f1fffb5aae  # main
     permissions:
       contents: write
       actions: read

.github/workflows/explosionbot.yml

@@ -6,6 +6,8 @@ on:
       - created
       - edited
 
+permissions: {}
+
 jobs:
   explosion-bot:
     if: github.repository_owner == 'explosion'
@@ -15,13 +17,15 @@ jobs:
         env:
           GITHUB_CONTEXT: ${{ toJson(github) }}
         run: echo "$GITHUB_CONTEXT"
-      - uses: actions/checkout@v4
-      - uses: actions/setup-python@v4
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd  # v6
+      - uses: actions/setup-python@a309ff8b426b58ec0e2a45f0f869d46889d02405  # v6
       - name: Install and run explosion-bot
         run: |
-          pip install git+https://${{ secrets.EXPLOSIONBOT_TOKEN }}@github.com/explosion/explosion-bot
+          git config --global url."https://x-access-token:${EXPLOSIONBOT_TOKEN}@github.com/".insteadOf "https://github.com/"
+          pip install git+https://github.com/explosion/explosion-bot
           python -m explosionbot
         env:
+          EXPLOSIONBOT_TOKEN: ${{ secrets.EXPLOSIONBOT_TOKEN }}
           INPUT_TOKEN: ${{ secrets.EXPLOSIONBOT_TOKEN }}
           INPUT_BK_TOKEN: ${{ secrets.BUILDKITE_SECRET }}
           ENABLED_COMMANDS: "test_gpu,test_slow,test_slow_gpu"

.github/workflows/issue-manager.yml

@@ -11,12 +11,16 @@ on:
     types:
       - labeled
 
+permissions: {}
+
 jobs:
   issue-manager:
+    permissions:
+      issues: write
     if: github.repository_owner == 'explosion'
     runs-on: ubuntu-latest
     steps:
-      - uses: tiangolo/issue-manager@0.4.0
+      - uses: tiangolo/issue-manager@4d1b7e05935a404dc8337d30bd23be46be8bb8e5  # 0.4.0
         with:
           token: ${{ secrets.GITHUB_TOKEN }}
           config: >

.github/workflows/lock.yml

@@ -16,7 +16,7 @@ jobs:
     if: github.repository_owner == 'explosion'
     runs-on: ubuntu-latest
     steps:
-      - uses: dessant/lock-threads@v5
+      - uses: dessant/lock-threads@1bf7ec25051fe7c00bdd17e6a7cf3d7bfb7dc771  # v5
         with:
           process-only: 'issues'
           issue-inactive-days: '30'

.github/workflows/publish_pypi.yml

@@ -8,6 +8,8 @@ on:
     types:
       - published
 
+permissions: {}
+
 jobs:
   upload_pypi:
     runs-on: ubuntu-latest
@@ -21,7 +23,7 @@ jobs:
     # or, alternatively, upload to PyPI on every tag starting with 'v' (remove on: release above to use this)
     # if: github.event_name == 'push' && startsWith(github.ref, 'refs/tags/v')
     steps:
-      - uses: robinraju/release-downloader@v1
+      - uses: robinraju/release-downloader@daf26c55d821e836577a15f77d86ddc078948b05  # v1
         with:
           tag: ${{ github.event.release.tag_name }}
           fileName: '*'

.github/workflows/spacy_universe_alert.yml

@@ -5,21 +5,16 @@ on:
     paths:
       - "website/meta/universe.json"
 
+permissions: {}
+
 jobs:
   build:
     if: github.repository_owner == 'explosion'
     runs-on: ubuntu-latest
 
     steps:
-      - name: Dump GitHub context
-        env:
-          GITHUB_CONTEXT: ${{ toJson(github) }}
-          PR_NUMBER: ${{github.event.number}}
-        run: |
-          echo "$GITHUB_CONTEXT"
-
-      - uses: actions/checkout@v4
-      - uses: actions/setup-python@v4
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd  # v6
+      - uses: actions/setup-python@a309ff8b426b58ec0e2a45f0f869d46889d02405  # v6
         with:
           python-version: '3.10'
       - name: Install Bernadette app dependency and send an alert

.github/workflows/tests.yml

@@ -19,6 +19,8 @@ on:
       - "*.mdx"
       - "website/**"
 
+permissions: {}
+
 jobs:
   validate:
     name: Validate
@@ -26,10 +28,10 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Check out repo
-        uses: actions/checkout@v4
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd  # v6
 
       - name: Configure Python version
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@a309ff8b426b58ec0e2a45f0f869d46889d02405  # v6
         with:
           python-version: "3.10"
 
@@ -45,19 +47,19 @@ jobs:
     name: Test
     needs: Validate
     strategy:
-      fail-fast: true
+      fail-fast: false
       matrix:
         os: [ubuntu-latest, windows-latest, macos-latest]
-        python_version: ["3.10", "3.11", "3.12", "3.13"]
+        python_version: ["3.10", "3.11", "3.12", "3.13", "3.14"]
 
     runs-on: ${{ matrix.os }}
 
     steps:
       - name: Check out repo
-        uses: actions/checkout@v4
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd  # v6
 
       - name: Configure Python version
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@a309ff8b426b58ec0e2a45f0f869d46889d02405  # v6
         with:
           python-version: ${{ matrix.python_version }}
 
@@ -93,7 +95,7 @@ jobs:
         shell: bash
 
       - name: Test import
-        run: python -W error -c "import spacy"
+        run: python -W error -W 'ignore:Core Pydantic V1:UserWarning:pydantic' -c "import spacy"
 
       - name: "Test download CLI"
         run: |
@@ -154,7 +156,7 @@ jobs:
 
       - name: "Run CPU tests"
         run: |
-          python -m pytest --pyargs spacy -W error
+          python -m pytest --pyargs spacy -W error -W 'ignore:Core Pydantic V1:UserWarning:pydantic'
         if: "!(startsWith(matrix.os, 'macos') && matrix.python_version == '3.11')"
 
       - name: "Run CPU tests with thinc-apple-ops"

.github/workflows/universe_validation.yml

@@ -13,6 +13,8 @@ on:
     paths:
       - "website/meta/universe.json"
 
+permissions: {}
+
 jobs:
   validate:
     name: Validate
@@ -20,10 +22,10 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Check out repo
-        uses: actions/checkout@v4
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd  # v6
 
       - name: Configure Python version
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@a309ff8b426b58ec0e2a45f0f869d46889d02405  # v6
         with:
           python-version: "3.7"
 

pyproject.toml

@@ -5,7 +5,7 @@ requires = [
     "cymem>=2.0.2,<2.1.0",
     "preshed>=3.0.2,<3.1.0",
     "murmurhash>=0.28.0,<1.1.0",
-    "thinc>=8.3.4,<8.4.0",
+    "thinc>=8.3.12,<8.4.0",
     "numpy>=2.0.0,<3.0.0"
 ]
 build-backend = "setuptools.build_meta"

requirements.txt

@@ -3,19 +3,19 @@ spacy-legacy>=3.0.11,<3.1.0
 spacy-loggers>=1.0.0,<2.0.0
 cymem>=2.0.2,<2.1.0
 preshed>=3.0.2,<3.1.0
-thinc>=8.3.4,<8.4.0
-ml_datasets>=0.2.0,<0.3.0
+thinc>=8.3.12,<8.4.0
+ml_datasets>=0.2.1,<0.3.0
 murmurhash>=0.28.0,<1.1.0
 wasabi>=0.9.1,<1.2.0
-srsly>=2.4.3,<3.0.0
+srsly>=2.5.3,<3.0.0
 catalogue>=2.0.6,<2.1.0
 typer>=0.3.0,<1.0.0
-weasel>=0.4.2,<0.5.0
+weasel>=1.0.0,<2.0.0
 # Third party dependencies
 numpy>=2.0.0,<3.0.0
 requests>=2.13.0,<3.0.0
 tqdm>=4.38.0,<5.0.0
-pydantic>=1.7.4,!=1.8,!=1.8.1,<3.0.0
+pydantic>=2.0.0,<3.0.0
 jinja2
 # Official Python utilities
 setuptools
@@ -34,4 +34,4 @@ types-requests
 types-setuptools>=57.0.0
 ruff>=0.9.0
 cython-lint>=0.15.0
-confection>=0.0.4,<1.0.0
+confection>=1.1.0,<2.0.0

setup.cfg

@@ -22,6 +22,7 @@ classifiers =
     Programming Language :: Python :: 3.11
     Programming Language :: Python :: 3.12
     Programming Language :: Python :: 3.13
+    Programming Language :: Python :: 3.14
     Topic :: Scientific/Engineering
 project_urls =
     Release notes = https://github.com/explosion/spaCy/releases
@@ -41,7 +42,7 @@ setup_requires =
     cymem>=2.0.2,<2.1.0
     preshed>=3.0.2,<3.1.0
     murmurhash>=0.28.0,<1.1.0
-    thinc>=8.3.4,<8.4.0
+    thinc>=8.3.12,<8.4.0
 install_requires =
     # Our libraries
     spacy-legacy>=3.0.11,<3.1.0
@@ -49,19 +50,19 @@ install_requires =
     murmurhash>=0.28.0,<1.1.0
     cymem>=2.0.2,<2.1.0
     preshed>=3.0.2,<3.1.0
-    thinc>=8.3.4,<8.4.0
+    thinc>=8.3.12,<8.4.0
     wasabi>=0.9.1,<1.2.0
-    srsly>=2.4.3,<3.0.0
+    srsly>=2.5.3,<3.0.0
     catalogue>=2.0.6,<2.1.0
-    weasel>=0.4.2,<0.5.0
-    confection>=0.0.4,<1.0.0
+    weasel>=1.0.0,<2.0.0
+    confection>=1.1.0,<2.0.0
     # Third-party dependencies
     typer>=0.3.0,<1.0.0
     tqdm>=4.38.0,<5.0.0
     numpy>=1.15.0; python_version < "3.9"
     numpy>=1.19.0; python_version >= "3.9"
     requests>=2.13.0,<3.0.0
-    pydantic>=1.7.4,!=1.8,!=1.8.1,<3.0.0
+    pydantic>=2.0.0,<3.0.0
     jinja2
     # Official Python utilities
     setuptools
@@ -135,6 +136,9 @@ formats = gztar
 markers =
     slow: mark a test as slow
     issue: reference specific issue
+filterwarnings =
+    error
+    ignore:Core Pydantic V1:UserWarning:pydantic
 
 [mypy]
 ignore_missing_imports = True

spacy/__init__.py

@@ -41,7 +41,7 @@ for _schema in (
     ConfigSchemaInit,
     ConfigSchema,
 ):
-    _schema.model_rebuild(_types_namespace=_rebuild_ns)
+    _schema.model_rebuild(_types_namespace=_rebuild_ns)  # type: ignore[attr-defined]
 
 if sys.maxunicode == 65535:
     raise SystemError(Errors.E130)

spacy/cli/debug_config.py

@@ -82,10 +82,10 @@ def debug_config(
         config = nlp.config.interpolate()
     msg.divider("Config validation for [initialize]")
     with show_validation_error(config_path):
-        T = registry.resolve(config["initialize"], schema=ConfigSchemaInit)
+        T = registry.resolve(config["initialize"], schema=ConfigSchemaInit)  # type: ignore[arg-type]
     msg.divider("Config validation for [training]")
     with show_validation_error(config_path):
-        T = registry.resolve(config["training"], schema=ConfigSchemaTraining)
+        T = registry.resolve(config["training"], schema=ConfigSchemaTraining)  # type: ignore[arg-type]
         dot_names = [T["train_corpus"], T["dev_corpus"]]
         util.resolve_dot_names(config, dot_names)
     msg.good("Config is valid")

spacy/cli/debug_data.py

@@ -137,7 +137,7 @@ def debug_data(
         cfg = util.load_config(config_path, overrides=config_overrides)
         nlp = util.load_model_from_config(cfg)
         config = nlp.config.interpolate()
-        T = registry.resolve(config["training"], schema=ConfigSchemaTraining)
+        T = registry.resolve(config["training"], schema=ConfigSchemaTraining)  # type: ignore[arg-type]
     # Use original config here, not resolved version
     sourced_components = get_sourced_components(cfg)
     frozen_components = T["frozen_components"]

spacy/cli/debug_model.py

@@ -89,7 +89,7 @@ def debug_model_cli(
     with show_validation_error(config_path):
         nlp = util.load_model_from_config(raw_config)
         config = nlp.config.interpolate()
-        T = registry.resolve(config["training"], schema=ConfigSchemaTraining)
+        T = registry.resolve(config["training"], schema=ConfigSchemaTraining)  # type: ignore[arg-type]
     seed = T["seed"]
     if seed is not None:
         msg.info(f"Fixing random seed: {seed}")

spacy/cli/find_threshold.py

@@ -207,8 +207,10 @@ def find_threshold(
             ),
         )
         if hasattr(pipe, "cfg"):
-            nlp.get_pipe(pipe_name).cfg = set_nested_item(
-                pipe.cfg, config_keys, threshold
+            nlp.get_pipe(pipe_name).cfg = set_nested_item(  # type: ignore[attr-defined]
+                pipe.cfg,
+                config_keys,
+                threshold,  # type: ignore[attr-defined]
             )
 
         eval_scores = nlp.evaluate(dev_dataset)

spacy/cli/init_config.py

@@ -216,7 +216,7 @@ def init_config(
     # Filter out duplicates since tok2vec and transformer are added by template
     pipeline = [pipe for pipe in pipeline if pipe not in ("tok2vec", "transformer")]
     defaults = RECOMMENDATIONS["__default__"]
-    reco = RecommendationSchema(**RECOMMENDATIONS.get(lang, defaults)).dict()
+    reco = RecommendationSchema(**RECOMMENDATIONS.get(lang, defaults)).model_dump()
     variables = {
         "lang": lang,
         "components": pipeline,

spacy/compat.py

@@ -35,8 +35,8 @@ else:
 try:  # Python 3.8+
     import importlib.metadata as importlib_metadata
 except ImportError:
-    from catalogue import (
-        _importlib_metadata as importlib_metadata,  # type: ignore[no-redef]    # noqa: F401
+    from catalogue import (  # type: ignore[no-redef]
+        _importlib_metadata as importlib_metadata,  # noqa: F401
     )
 
 from thinc.api import Optimizer  # noqa: F401

spacy/lang/es/lemmatizer.py

@@ -416,7 +416,7 @@ class SpanishLemmatizer(Lemmatizer):
             rule = self.select_rule("verb", features)
             verb_lemma = self.lemmatize_verb(
                 verb,
-                features - {"PronType=Prs"},
+                features - {"PronType=Prs"},  # type: ignore[operator]
                 rule,
                 index,  # type: ignore[operator]
             )[0]

spacy/language.py

@@ -1323,7 +1323,7 @@ class Language:
         # Make sure the config is interpolated so we can resolve subsections
         config = self.config.interpolate()
         # These are the settings provided in the [initialize] block in the config
-        I = registry.resolve(config["initialize"], schema=ConfigSchemaInit)
+        I = registry.resolve(config["initialize"], schema=ConfigSchemaInit)  # type: ignore[arg-type]
         before_init = I["before_init"]
         if before_init is not None:
             before_init(self)
@@ -1353,7 +1353,7 @@ class Language:
                 proc.initialize(get_examples, nlp=self, **p_settings)
         pretrain_cfg = config.get("pretraining")
         if pretrain_cfg:
-            P = registry.resolve(pretrain_cfg, schema=ConfigSchemaPretrain)
+            P = registry.resolve(pretrain_cfg, schema=ConfigSchemaPretrain)  # type: ignore[arg-type]
             init_tok2vec(self, P, I)
         self._link_components()
         self._optimizer = sgd
@@ -1823,7 +1823,7 @@ class Language:
         orig_pretraining = config.pop("pretraining", None)
         config["components"] = {}
         if auto_fill:
-            filled = registry.fill(config, validate=validate, schema=ConfigSchema)
+            filled = registry.fill(config, validate=validate, schema=ConfigSchema)  # type: ignore[arg-type]
         else:
             filled = config
         filled["components"] = orig_pipeline
@@ -1832,7 +1832,9 @@ class Language:
             filled["pretraining"] = orig_pretraining
             config["pretraining"] = orig_pretraining
         resolved_nlp = registry.resolve(
-            filled["nlp"], validate=validate, schema=ConfigSchemaNlp
+            filled["nlp"],
+            validate=validate,
+            schema=ConfigSchemaNlp,  # type: ignore[arg-type]
         )
         create_tokenizer = resolved_nlp["tokenizer"]
         create_vectors = resolved_nlp["vectors"]

spacy/matcher/phrasematcher.pyx

@@ -57,7 +57,7 @@ cdef class PhraseMatcher:
                 attr = "ORTH"
             if attr == "IS_SENT_START":
                 attr = "SENT_START"
-            if attr.lower() not in TokenPattern().dict():
+            if attr.lower() not in TokenPattern().model_dump():
                 raise ValueError(Errors.E152.format(attr=attr))
             self.attr = IDS.get(attr)
 

spacy/pipeline/_edit_tree_internals/schemas.py

@@ -6,6 +6,7 @@ from pydantic import (
     ConfigDict,
     Field,
     RootModel,
+    StrictBool,
     StrictInt,
     StrictStr,
     ValidationError,

spacy/pipeline/factories.py

@@ -14,9 +14,10 @@ from ..pipeline.edit_tree_lemmatizer import (
 )
 
 # Import factory default configurations
-from ..pipeline.entity_linker import DEFAULT_NEL_MODEL, EntityLinker, EntityLinker_v1
+from ..pipeline.entity_linker import DEFAULT_NEL_MODEL, EntityLinker
 from ..pipeline.entityruler import DEFAULT_ENT_ID_SEP, EntityRuler
 from ..pipeline.functions import DocCleaner, TokenSplitter
+from ..pipeline.legacy import EntityLinker_v1
 from ..pipeline.lemmatizer import Lemmatizer
 from ..pipeline.morphologizer import DEFAULT_MORPH_MODEL, Morphologizer
 from ..pipeline.multitask import DEFAULT_MT_MODEL, MultitaskObjective

spacy/schemas.py

@@ -251,7 +251,7 @@ class TokenPatternOperatorSimple(str, Enum):
 TokenPatternOperatorMinMax = constr(pattern=r"^(\{\d+\}|\{\d+,\d*\}|\{\d*,\d+\})$")
 
 
-TokenPatternOperator = Union[TokenPatternOperatorSimple, TokenPatternOperatorMinMax]
+TokenPatternOperator = Union[TokenPatternOperatorSimple, TokenPatternOperatorMinMax]  # type: ignore[valid-type]
 StringValue = Union[TokenPatternString, StrictStr]
 NumberValue = Union[TokenPatternNumber, StrictInt, StrictFloat]
 UnderscoreValue = Union[
@@ -420,8 +420,8 @@ class ConfigSchemaInit(BaseModel):
     lookups: Optional[Lookups] = Field(..., title="Vocabulary lookups, e.g. lexeme normalization")
     vectors: Optional[StrictStr] = Field(..., title="Path to vectors")
     init_tok2vec: Optional[StrictStr] = Field(..., title="Path to pretrained tok2vec weights")
-    tokenizer: Dict[StrictStr, Any] = Field(..., help="Arguments to be passed into Tokenizer.initialize")
-    components: Dict[StrictStr, Dict[StrictStr, Any]] = Field(..., help="Arguments for TrainablePipe.initialize methods of pipeline components, keyed by component")
+    tokenizer: Dict[StrictStr, Any] = Field(..., title="Arguments to be passed into Tokenizer.initialize")
+    components: Dict[StrictStr, Dict[StrictStr, Any]] = Field(..., title="Arguments for TrainablePipe.initialize methods of pipeline components, keyed by component")
     before_init: Optional[Callable[["Language"], "Language"]] = Field(..., title="Optional callback to modify nlp object before initialization")
     after_init: Optional[Callable[["Language"], "Language"]] = Field(..., title="Optional callback to modify nlp object after initialization")
     # fmt: on

spacy/tests/lang/zh/test_tokenizer.py

@@ -1,5 +1,5 @@
 import pytest
-from thinc.api import ConfigValidationError
+from confection import ConfigValidationError
 
 from spacy.lang.zh import Chinese, _get_pkuseg_trie_data
 

spacy/tests/package/test_requirements.py

@@ -15,6 +15,7 @@ def test_build_dependencies():
         "cython-lint",
         "black",
         "isort",
+        "ruff",
         "mypy",
         "types-dataclasses",
         "types-mock",

spacy/tests/pipeline/test_textcat.py

@@ -24,6 +24,7 @@ from spacy.pipeline.textcat_multilabel import (
 )
 from spacy.pipeline.tok2vec import DEFAULT_TOK2VEC_MODEL
 from spacy.scorer import Scorer
+from spacy.tests.tok2vec import build_lazy_init_tok2vec as _  # noqa: F401
 from spacy.tokens import Doc, DocBin
 from spacy.training import Example
 from spacy.training.initialize import init_nlp

spacy/tests/test_cli.py

@@ -1229,6 +1229,9 @@ def test_download_rejects_relative_urls(monkeypatch):
     relative path in the filename"""
 
     monkeypatch.setattr(download_module, "run_command", lambda cmd: None)
+    monkeypatch.setattr(
+        download_module, "_get_pip_install_cmd", lambda: ["pip", "install"]
+    )
 
     # Check that normal download works
     download_module.download("en_core_web_sm-3.7.1", direct=True)

spacy/training/batchers.py

@@ -4,6 +4,7 @@ from typing import (
     Any,
     Callable,
     Iterable,
+    Iterator,
     List,
     Optional,
     Sequence,

spacy/training/corpus.py

@@ -50,9 +50,10 @@ def create_jsonl_reader(
 
 
 @util.registry.readers("spacy.read_labels.v1")
-def read_labels(path: Path, *, require: bool = False):
+def read_labels(path: Union[str, Path], *, require: bool = False):
     # I decided not to give this a generic name, because I don't want people to
     # use it for arbitrary stuff, as I want this require arg with default False.
+    path = Path(path)
     if not require and not path.exists():
         return None
     return srsly.read_json(path)

spacy/training/initialize.py

@@ -50,7 +50,7 @@ def init_nlp(config: Config, *, use_gpu: int = -1) -> "Language":
     logger.info("Set up nlp object from config")
     config = nlp.config.interpolate()
     # Resolve all training-relevant sections using the filled nlp config
-    T = registry.resolve(config["training"], schema=ConfigSchemaTraining)
+    T = registry.resolve(config["training"], schema=ConfigSchemaTraining)  # type: ignore[arg-type]
     dot_names = [T["train_corpus"], T["dev_corpus"]]
     if not isinstance(T["train_corpus"], str):
         raise ConfigValidationError(

spacy/training/loop.py

@@ -62,7 +62,7 @@ def train(
     allocator = config["training"]["gpu_allocator"]
     if use_gpu >= 0 and allocator:
         set_gpu_allocator(allocator)
-    T = registry.resolve(config["training"], schema=ConfigSchemaTraining)
+    T = registry.resolve(config["training"], schema=ConfigSchemaTraining)  # type: ignore[arg-type]
     dot_names = [T["train_corpus"], T["dev_corpus"]]
     train_corpus, dev_corpus = resolve_dot_names(config, dot_names)
     optimizer = T["optimizer"]

spacy/training/pretrain.py

@@ -42,7 +42,7 @@ def pretrain(
     config["initialize"]["init_tok2vec"] = None
     nlp = load_model_from_config(config)
     _config = nlp.config.interpolate()
-    P = registry.resolve(_config["pretraining"], schema=ConfigSchemaPretrain)
+    P = registry.resolve(_config["pretraining"], schema=ConfigSchemaPretrain)  # type: ignore[arg-type]
     corpus = dot_to_object(_config, P["corpus"])
     corpus = registry.resolve({"corpus": corpus})["corpus"]
     batcher = P["batcher"]