Feat/hashing vectorizer (#268)

* add embedder

* add test embedder config

* update tests

* update assertions in some tests

* Feat/extras matrix for tests (#269)

* add dependencies matrix

* add pipeline tests for each scoring module

* bug fix

* upd extras installation in gh actions

* upd extras in mypy ci

* fix issues with dependencies for bert scorer test

* fix gcn scorer tests

* fix typing errors

* upd assertions about catboost predictions

* upd ci with presets tests

* try to fix unit tests

* Feat/extras matrix for tests (#270)

* add dependencies matrix

* add pipeline tests for each scoring module

* bug fix

* upd extras installation in gh actions

* upd extras in mypy ci

* fix issues with dependencies for bert scorer test

* fix gcn scorer tests

* fix typing errors

* upd assertions about catboost predictions

* upd ci with presets tests

* try to fix unit tests

* try to fix mypy

* fix catboost test

* skip incremental evolver tests for now

* upd callback test

* run ruff

* add missing transformers dependency to embedder ci

* update test for tunable decision module

* move callback testing to the env with sentence-transformers installed

* fix embedder tests

* remove catboost from classic-medium

* lint

* small fixes

* lint

---------

Co-authored-by: Roman Solomatin <36135455+Samoed@users.noreply.github.com>

Author: voorhs

.github/workflows/reusable-test.yaml

@@ -7,6 +7,11 @@ on:
         required: true
         type: string
         description: 'Command to run tests'
+      extras:
+        required: false
+        type: string
+        default: ''
+        description: 'Space-separated --extra flags (e.g., "--extra transformers --extra peft")'
 
 jobs:
   test:
@@ -15,7 +20,7 @@ jobs:
       fail-fast: false
       matrix:
         os: [ ubuntu-latest ]
-        python-version: [ "3.10", "3.11", "3.12", "3.13", ]
+        python-version: [ "3.10", "3.11", "3.12", "3.13", "3.14"]
         include:
           - os: windows-latest
             python-version: "3.10"
@@ -40,7 +45,7 @@ jobs:
     - name: Install dependencies for Python ${{ matrix.python-version }}
       run: |
         uv python pin ${{ matrix.python-version }}
-        uv sync --group test --extra catboost --extra peft --extra transformers --extra sentence-transformers --extra openai
+        uv sync --group test ${{ inputs.extras }}
 
     - name: Run tests
       run: |

.github/workflows/test-embedder.yaml

@@ -0,0 +1,15 @@
+name: test embedder
+
+on:
+  push:
+    branches:
+      - dev
+  pull_request:
+
+jobs:
+  test:
+    uses: ./.github/workflows/reusable-test.yaml
+    with:
+      test_command: pytest -n auto tests/embedder/ tests/callback/
+      extras: --extra sentence-transformers --extra transformers
+

.github/workflows/test-inference.yaml

@@ -11,3 +11,4 @@ jobs:
     uses: ./.github/workflows/reusable-test.yaml
     with:
       test_command: pytest -n auto tests/pipeline/test_inference.py
+      extras: --extra catboost --extra peft --extra transformers --extra sentence-transformers

.github/workflows/test-optimization.yaml

@@ -11,3 +11,4 @@ jobs:
     uses: ./.github/workflows/reusable-test.yaml
     with:
       test_command: pytest -n auto tests/pipeline/test_optimization.py
+      extras: --extra catboost --extra peft --extra transformers --extra sentence-transformers

.github/workflows/test-presets.yaml

@@ -11,3 +11,4 @@ jobs:
     uses: ./.github/workflows/reusable-test.yaml
     with:
       test_command: pytest -n auto tests/pipeline/test_presets.py
+      extras: --extra catboost --extra peft --extra transformers --extra sentence-transformers

.github/workflows/test-scorers.yaml

@@ -0,0 +1,47 @@
+name: test scorers
+
+on:
+  push:
+    branches:
+      - dev
+  pull_request:
+
+jobs:
+  test:
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ ubuntu-latest ]
+        python-version: [ "3.10", "3.11", "3.12" ]
+        dependency-group: [ "base", "transformers", "peft", "catboost" ]
+        include:
+          - os: windows-latest
+            python-version: "3.10"
+            dependency-group: "base"
+
+    steps:
+    - name: Checkout code
+      uses: actions/checkout@v4
+
+    - name: Cache Hugging Face
+      id: cache-hf
+      uses: actions/cache@v4
+      with:
+        path: ~/.cache/huggingface
+        key: ${{ runner.os }}-hf
+
+    - name: Install uv
+      uses: astral-sh/setup-uv@v6
+      with:
+        version: "0.8.8"
+
+    - name: Install dependencies for Python ${{ matrix.python-version }}
+      run: |
+        uv python pin ${{ matrix.python-version }}
+        uv sync --group test ${{ matrix.dependency-group != 'base' && format('--extra {0}', matrix.dependency-group) || '' }}
+
+    - name: Run scorer tests
+      run: |
+        uv run pytest -n auto tests/modules/scoring/
+

.github/workflows/unit-tests.yaml

@@ -10,4 +10,4 @@ jobs:
   test:
     uses: ./.github/workflows/reusable-test.yaml
     with:
-      test_command: pytest -n auto --ignore=tests/nodes --ignore=tests/pipeline
+      test_command: pytest -n auto --ignore=tests/modules/scoring/ --ignore=tests/pipeline --ignore=tests/embedder --ignore=tests/callback

src/autointent/_dump_tools/unit_dumpers.py

@@ -292,7 +292,7 @@ def load(path: Path, **kwargs: Any) -> PreTrainedTokenizer | PreTrainedTokenizer
         require("transformers", extra="transformers")
         import transformers
 
-        return transformers.AutoTokenizer.from_pretrained(path)
+        return transformers.AutoTokenizer.from_pretrained(path)  # type: ignore[no-any-return,no-untyped-call]
 
     @classmethod
     def check_isinstance(cls, obj: Any) -> bool:  # noqa: ANN401

src/autointent/_presets/classic-medium.yaml

@@ -12,7 +12,6 @@ search_space:
         k:
           low: 1
           high: 20
-      - module_name: catboost
       - module_name: sklearn
         clf_name: [RandomForestClassifier]
         n_estimators: [150]

src/autointent/_wrappers/embedder/__init__.py

@@ -2,12 +2,14 @@
 
 from .base import BaseEmbeddingBackend
 from .embedder import Embedder
+from .hashing_vectorizer import HashingVectorizerEmbeddingBackend
 from .openai import OpenaiEmbeddingBackend
 from .sentence_transformers import SentenceTransformerEmbeddingBackend
 
 __all__ = [
     "BaseEmbeddingBackend",
     "Embedder",
+    "HashingVectorizerEmbeddingBackend",
     "OpenaiEmbeddingBackend",
     "SentenceTransformerEmbeddingBackend",
 ]