Release 1.0.0: add example

.github/workflows/ci.yml

@@ -28,4 +28,37 @@ jobs:
 
       - name: Run tests
         run: |
-          poetry run pytest test_integration.py
+          make test
+
+      - uses: actions-ecosystem/action-regex-match@v2
+        id: regex-match
+        with:
+          text: ${{ github.event.head_commit.message }}
+          regex: "^Release ([^ ]+)"
+
+      - name: Build package
+        run: |
+          poetry build
+
+      - name: Publish to PyPI
+        if: ${{ steps.regex-match.outputs.match != '' }}
+        run: |
+          poetry config pypi-token.pypi ${{ secrets.PYPI_API_TOKEN }}
+          poetry publish
+
+      - uses: rickstaa/action-create-tag@v1
+        if: ${{ steps.regex-match.outputs.match != '' }}
+        id: "tag_create"
+        with:
+          tag: ${{ env.version }}
+          tag_exists_error: true
+          message: "Release ${{ steps.regex-match.outputs.group1 }}"
+
+      - name: Create Release
+        if: ${{ steps.regex-match.outputs.match != '' }}
+        uses: softprops/action-gh-release@v1
+        with:
+          tag_name: ${{ env.version }}
+          name: ${{ env.version }}
+          draft: false
+          prerelease: false

Makefile

@@ -1,3 +1,15 @@
+VERSION="$(shell python3 -c 'import toml; print(toml.load("pyproject.toml")["tool"]["poetry"]["version"])')"
+
+#* Release
+
+release: ## [Local development] Release a new version of the API.
+	@echo "Releasing version ${VERSION}"; \
+	read -p "Commit content:" COMMIT; \
+	git add .; \
+	echo "Committing '${VERSION}: $$COMMIT'"; \
+	git commit -m "Release ${VERSION}: $$COMMIT"; \
+	git push origin main
+	@echo "Done, check '\033[0;31mhttps://github.com/different-ai/embedbase-qdrant/actions\033[0m'"
 
 #* Testing
 

examples/main.py

@@ -0,0 +1,45 @@
+from embedbase import get_app
+from embedbase.embedding.base import Embedder
+import uvicorn
+from sentence_transformers import SentenceTransformer
+from embedbase_qdrant import Qdrant
+
+class LocalEmbedder(Embedder):
+    EMBEDDING_MODEL = "all-MiniLM-L6-v2"
+
+    def __init__(self, model: str = EMBEDDING_MODEL, **kwargs):
+        super().__init__(**kwargs)
+        self.model = SentenceTransformer(model)
+        self._dimensions = self.model.get_sentence_embedding_dimension()
+
+    @property
+    def dimensions(self) -> int:
+        """
+        Return the dimensions of the embeddings
+        :return: dimensions of the embeddings
+        """
+        return self._dimensions
+
+    def is_too_big(self, text: str) -> bool:
+        """
+        Check if text is too big to be embedded,
+        delegating the splitting UX to the caller
+        :param text: text to check
+        :return: True if text is too big, False otherwise
+        """
+        return len(text) > self.model.get_max_seq_length()
+
+    async def embed(self, data):
+        """
+        Embed a list of strings or a single string
+        :param data: list of strings or a single string
+        :return: list of embeddings
+        """
+        embeddings = self.model.encode(data)
+        return embeddings.tolist() if isinstance(data, list) else [embeddings.tolist()]
+
+
+app = get_app().use_embedder(LocalEmbedder()).use_db(Qdrant()).run()
+
+if __name__ == "__main__":
+    uvicorn.run(app, reload=True)

examples/requirements.txt

@@ -0,0 +1,3 @@
+embedbase
+sentence-transformers
+git+https://github.com/different-ai/embedbase-qdrant.git