run

Dockerfile.base

@@ -8,7 +8,7 @@ RUN mkdir -p /opt/osm
 RUN pip install uv fastapi[standard] uvicorn
 
 # LLM stuff
-RUN pip install llama-index llama-index-llms-openai llama-index-program-openai
+RUN pip install llama-index llama-index-llms-openai llama-index-program-openai llama-index-llms-openrouter
 
 # Dashboard stuff
 RUN pip install colorcet panel

compose.development.override.yaml

@@ -1,56 +1,57 @@
 name: local-osm
 services:
-  rtransparent:
-    container_name: rtransparent
-    build:
-      context: .
-      dockerfile: ./external_components/rtransparent/Dockerfile
-    volumes:
-      - ./external_components/rtransparent:/app
+  # rtransparent:
+  #   container_name: rtransparent
+  #   build:
+  #     context: .
+  #     dockerfile: ./external_components/rtransparent/Dockerfile
+  #   volumes:
+  #     - ./external_components/rtransparent:/app
 
   llm_extraction:
     container_name: llm_extraction
     environment:
       - OPENAI_API_KEY=${OPENAI_API_KEY:-NOKEY}
+      - OPENROUTER_API_KEY=${OPENROUTER_API_KEY:-NOKEY}
     build:
       context: .
       dockerfile: ./external_components/llm_extraction/Dockerfile
     volumes:
       - ./external_components/llm_extraction:/app
 
-  web_api:
-    container_name: web_api
-    environment:
-      - MONGODB_URI=mongodb://db:27017/osm
-    build:
-      context: .
-      dockerfile: ./web/api/Dockerfile
-    ports:
-      - 80:80
-    volumes:
-      - ./web/api:/app/app
-      - ./osm:/opt/osm/osm
-    working_dir: /app/app
-    command: ["fastapi","dev","--host","0.0.0.0","--port","80"]
-    depends_on:
-      - db
+  # web_api:
+  #   container_name: web_api
+  #   environment:
+  #     - MONGODB_URI=mongodb://db:27017/osm
+  #   build:
+  #     context: .
+  #     dockerfile: ./web/api/Dockerfile
+  #   ports:
+  #     - 80:80
+  #   volumes:
+  #     - ./web/api:/app/app
+  #     - ./osm:/opt/osm/osm
+  #   working_dir: /app/app
+  #   command: ["fastapi","dev","--host","0.0.0.0","--port","80"]
+  #   depends_on:
+  #     - db
 
-  dashboard:
-    container_name: dashboard
-    environment:
-      - MONGODB_URI=mongodb://db:27017/osm
-    build:
-      context: .
-      dockerfile: ./web/dashboard/Dockerfile
-    ports:
-      - "8501:8501"
-    volumes:
-      - ./web/dashboard:/app
-      - ./osm:/opt/osm/osm
-    working_dir: /app
-    command: ["python", "app.py"]
-    depends_on:
-      - db
+  # dashboard:
+  #   container_name: dashboard
+  #   environment:
+  #     - MONGODB_URI=mongodb://db:27017/osm
+  #   build:
+  #     context: .
+  #     dockerfile: ./web/dashboard/Dockerfile
+  #   ports:
+  #     - "8501:8501"
+  #   volumes:
+  #     - ./web/dashboard:/app
+  #     - ./osm:/opt/osm/osm
+  #   working_dir: /app
+  #   command: ["python", "app.py"]
+  #   depends_on:
+  #     - db
 
   db:
     container_name: db

compose.yaml

@@ -4,11 +4,11 @@ services:
     image: elifesciences/sciencebeam-parser
     ports:
       - "8070:8070"
-  rtransparent:
-    container_name: rtransparent
-    image: nimhdsst/rtransparent:staging
-    ports:
-      - "8071:8071"
+  # rtransparent:
+  #   container_name: rtransparent
+  #   image: nimhdsst/rtransparent:staging
+  #   ports:
+  #     - "8071:8071"
   llm_extraction:
     container_name: llm_extraction
     image: nimhdsst/llm_extraction:staging

external_components/llm_extraction/app.py

@@ -1,4 +1,5 @@
 import logging
+import os
 
 from fastapi import FastAPI, File, HTTPException, Query, UploadFile
 from llama_index.core import ChatPromptTemplate
@@ -9,8 +10,14 @@
 
 # from pydantic import BaseModel, Field
 from osm.schemas.metrics_schemas import LLMExtractorMetrics
-
-LLM_MODELS = {"gpt-4o-2024-08-06": OpenAI(model="gpt-4o-2024-08-06")}
+from llama_index.llms.openrouter import OpenRouter
+LLM_MODELS = {
+    # "gpt-4o-2024-08-06": OpenRouter(
+    # api_key=os.environ["OPENROUTER_API_KEY"],
+    # model="openai/chatgpt-4o-latest",
+    # )
+    "gpt-4o-2024-08-06": OpenAI(model="gpt-4o-2024-08-06")
+}
 
 
 logger = logging.getLogger(__name__)

osm/pipeline/core.py

@@ -110,11 +110,11 @@ def run(self, user_managed_compose: bool = False, llm_model: str = None):
                 extracted_metrics = extractor.run(
                     parsed_data, parser=parser.name, llm_model=llm_model
                 )
-                self.savers.save_osm(
-                    data=self.file_data,
-                    metrics=extracted_metrics,
-                    components=[parser, extractor, *self.savers],
-                )
+                # self.savers.save_osm(
+                #     data=self.file_data,
+                #     metrics=extracted_metrics,
+                #     components=[parser, extractor, *self.savers],
+                # )
                 self.savers.save_json(extracted_metrics, self.metrics_path)
 
     @staticmethod

osm/pipeline/savers.py

@@ -85,7 +85,6 @@ def _run(self, data: bytes, metrics: dict, components: list[schemas.Component]):
         print(f"Using OSM API: {osm_api}")
         # Build the payload
         schemas = get_metrics_schemas()
-        breakpoint()
         try:
             payload = {
                 "osm_version": __version__,