fixed imports and back to multistaging

Dockerfile

@@ -1,37 +1,38 @@
-# FROM python:3.9 as core
-#
-# COPY ./core/requirements.txt ./requirements.txt
-# RUN pip install -r requirements.txt
-#
-# WORKDIR /app
-#
-# COPY ./core .
-# RUN python ./initialization.py
-#
-# FROM python:3.9
-#
-# COPY ./api/requirements.txt ./requirements.txt
-# RUN pip install -r requirements.txt
-#
-# WORKDIR /app
-# COPY ./api .
-#
-# COPY --from=core /app/engine.pickle /app/engine.pickle
-#
-# EXPOSE 9999
-# ENTRYPOINT ["python", "service_manager.py"]
-FROM python:3.9 as build
+FROM python:3.9 as core
 
-COPY ./requirements.txt ./requirements.txt
+COPY ./core/requirements.txt ./requirements.txt
 RUN pip install -r requirements.txt
 
 WORKDIR /app
 
-COPY . .
+COPY ./core .
+RUN python ./initialization.py
 
-RUN mkdir -p /app/cache && chmod -R 777 /app/cache
+FROM python:3.9
 
-ENV TRANSFORMERS_CACHE=/app/cache
+COPY ./api/requirements.txt ./requirements.txt
+RUN pip install -r requirements.txt
+
+WORKDIR /app
+COPY ./api .
+
+COPY --from=core /app/engine.pickle /app/engine.pickle
 
 EXPOSE 9999
-ENTRYPOINT ["python", "run.py"]
+ENTRYPOINT ["python", "service_manager.py"]
+
+# FROM python:3.9 as build
+#
+# COPY ./requirements.txt ./requirements.txt
+# RUN pip install -r requirements.txt
+#
+# WORKDIR /app
+#
+# COPY . .
+#
+# RUN mkdir -p /app/cache && chmod -R 777 /app/cache
+#
+# ENV TRANSFORMERS_CACHE=/app/cache
+#
+# EXPOSE 9999
+# ENTRYPOINT ["python", "run.py"]

api/service_manager.py

@@ -1,4 +1,4 @@
-from web_server import app
+from .web_server import app
 import uvicorn
 
 
@@ -9,5 +9,4 @@ def run():
 
     uvicorn.run(app, host="0.0.0.0", port=8000)
 
-run()
 

api/web_server.py

@@ -1,7 +1,7 @@
 import dill
 from fastapi import FastAPI, HTTPException
 from pydantic import BaseModel
-from core.search_engine import PromptSearchEngine
+# from ..core.search_engine import PromptSearchEngine
 
 
 class Query(BaseModel):
@@ -11,7 +11,7 @@ class Query(BaseModel):
 
 app = FastAPI()
 
-with open('../core/engine.pickle', 'rb') as file:
+with open('./engine.pickle', 'rb') as file:
     serialized_engine = file.read()
 
 prompt_search_engine = dill.loads(serialized_engine)
@@ -19,8 +19,21 @@ prompt_search_engine = dill.loads(serialized_engine)
 
 @app.post("/search/")
 async def search(query: Query):
+    """
+    TODO
+    """
+
     try:
+        if not isinstance(query.prompt, str):
+            raise ValueError("Prompt must be a string")
+
+        if not isinstance(query.n, int):
+            raise ValueError("Prompt must be a string")
+
         results = prompt_search_engine.most_similar(query.prompt, query.n)
-        return {"results": results}
+        formatted_results = [{"score": float(score), "description": desc} for score, desc in results]
+
+        return formatted_results
+
     except Exception as e:
         raise HTTPException(status_code=500, detail=str(e))

core/initialization.py

@@ -1,6 +1,6 @@
 import dill
-from core.data.dataset import PromptDataset
-from core.search_engine import PromptSearchEngine
+from .search_engine import PromptSearchEngine
+from .data.dataset import PromptDataset
 
 
 def run():
@@ -18,4 +18,4 @@ def run():
     with open("engine.pickle", "wb") as file:
         file.write(serialized_engine)
 
-run()
+run()

core/requirements.txt

@@ -1,4 +1,5 @@
 datasets==2.20.0
 dill==0.3.8
 faiss_cpu==1.8.0.post1
+numpy==1.23.5
 sentence_transformers==3.0.1

core/search_engine.py

@@ -1,7 +1,7 @@
 from typing import List, Sequence, Tuple
 import numpy as np
 import faiss
-from core.vectorizer import Vectorizer
+from .vectorizer import Vectorizer
 
 
 class PromptSearchEngine(object):

requirements.txt

@@ -5,3 +5,4 @@ fastapi==0.111.1
 pydantic==2.8.2
 sentence_transformers==3.0.1
 uvicorn==0.30.3
+numpy==1.23.5