feat: initial commit

Dockerfile

@@ -0,0 +1,13 @@
+FROM python:3.10
+
+RUN useradd -m -u 1000 user
+USER user
+ENV PATH="/home/user/.local/bin:$PATH"
+
+WORKDIR /app
+
+COPY --chown=user ./requirements.txt requirements.txt
+RUN pip install --no-cache-dir --upgrade -r requirements.txt
+
+COPY --chown=user . /app
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

README.md

@@ -1,11 +1,10 @@
 ---
 title: Movie Search
-emoji: 🌖
-colorFrom: gray
-colorTo: blue
+emoji: 🎥
+colorFrom: red
+colorTo: pink
 sdk: docker
 pinned: false
-license: mit
 short_description: API for retrieving movie information
 ---
 

app.py

@@ -0,0 +1,68 @@
+import numpy as np
+from fastapi import FastAPI
+from pydantic import BaseModel
+from haystack import Pipeline
+from haystack.utils import ComponentDevice
+from haystack.components.joiners import DocumentJoiner
+from haystack.components.rankers import TransformersSimilarityRanker
+from haystack.document_stores.in_memory import InMemoryDocumentStore
+from haystack.components.embedders import SentenceTransformersTextEmbedder
+from haystack.components.retrievers.in_memory import InMemoryBM25Retriever, InMemoryEmbeddingRetriever
+
+
+app = FastAPI()
+
+
+class MovieRequest(BaseModel):
+    query: str
+
+
+# Setting up the pipeline components
+print("Loading documents...")
+document_store = InMemoryDocumentStore.load_from_disk("movie_document_store_short.json")
+print("Documents loaded!")
+
+text_embedder = SentenceTransformersTextEmbedder(
+    model="BAAI/bge-small-en-v1.5", device=ComponentDevice.from_str("cpu")
+)
+embedding_retriever = InMemoryEmbeddingRetriever(document_store)
+bm25_retriever = InMemoryBM25Retriever(document_store)
+
+document_joiner = DocumentJoiner()
+
+ranker = TransformersSimilarityRanker(model="BAAI/bge-reranker-base")
+
+
+# Creating the hybrid search pipeline
+hybrid_search = Pipeline()
+hybrid_search.add_component("text_embedder", text_embedder)
+hybrid_search.add_component("embedding_retriever", embedding_retriever)
+hybrid_search.add_component("bm25_retriever", bm25_retriever)
+hybrid_search.add_component("document_joiner", document_joiner)
+hybrid_search.add_component("ranker", ranker)
+
+hybrid_search.connect("text_embedder", "embedding_retriever")
+hybrid_search.connect("bm25_retriever", "document_joiner")
+hybrid_search.connect("embedding_retriever", "document_joiner")
+hybrid_search.connect("document_joiner", "ranker")
+
+
+@app.post("/retrieve_movie_info")
+def retrieve_movie_info(movie_request: MovieRequest):
+    result = hybrid_search.run(
+        {
+            "text_embedder": {"text": movie_request.query},
+            "bm25_retriever": {"query": movie_request.query},
+            "ranker": {"query": movie_request.query}
+        }
+    )
+    
+    out = [
+        {
+            "info": doc.meta,
+            "score": doc.score
+        }
+        for doc in result["ranker"]["documents"]
+    ]
+
+    return out

requirements.txt

@@ -0,0 +1,6 @@
+fastapi
+uvicorn[standard]
+transformers==4.46.3
+haystack-ai==2.5.0
+sentence-transformers==3.0.1
+accelerate==1.6.0