Feature/fix broken build (#286)

* fix pip command quotes

* fix pip command quotes

* fix default

* tweaks

* fix broken docker build

* fix the docker build

Dockerfile

@@ -11,7 +11,7 @@ COPY pyproject.toml poetry.lock* /app/
 RUN pip install poetry keyring
 
 # Export dependencies to requirements.txt
-RUN poetry export --without-hashes --format=requirements.txt --output=requirements.txt -E eval -E parsing
+RUN poetry export --without-hashes --format=requirements.txt --output=requirements.txt -E eval -E sentence_transformers
 
 # Install dependencies from requirements.txt
 RUN pip install --no-cache-dir --upgrade -r requirements.txt
@@ -22,6 +22,11 @@ RUN pip install gunicorn uvicorn
 # Copy the rest of the application code
 COPY . /app
 
+# Set the default configuration option
+ENV CONFIG_OPTION=default
+
+# Expose the port
 EXPOSE 8000
 
-CMD ["gunicorn", "r2r.examples.servers.basic_pipeline:app", "--bind", "0.0.0.0:8000", "--workers", "2", "--threads", "8", "--timeout", "0", "--worker-class", "uvicorn.workers.UvicornWorker"]
+# Set the command to run the application with Gunicorn
+CMD ["sh", "-c", "gunicorn \"r2r.examples.servers.basic_pipeline:create_app('$CONFIG_OPTION')\" --bind 0.0.0.0:8000 --workers 2 --threads 8 --timeout 0 --worker-class uvicorn.workers.UvicornWorker"]

README.md

@@ -48,7 +48,7 @@ Note - the example above uses [SciPhi Cloud](https://app.sciphi.ai) to pair with
 
 ```bash
 # use the `'r2r[all]'` to download all required deps
-pip install 'r2r[parsing,eval]'
+pip install 'r2r[eval]'
 
 # setup env 
 export OPENAI_API_KEY=sk-...
@@ -65,8 +65,11 @@ export LOCAL_DB_PATH=local.sqlite
 ```bash
 docker pull emrgntcmplxty/r2r:latest
 
-# Place your secrets in `.env`
-docker run -d --name r2r_container -p 8000:8000 --env-file .env r2r
+# Choose from CONFIG_OPTION in {`default`, `local_ollama`}
+# For cloud deployment, select `default` and place your secrets in `.env`
+# For local deployment, select `local_ollama`
+docker run -d --name r2r_container -p 8000:8000 -e CONFIG_OPTION=local_ollama --env-file .env emrgntcmplxty/r2r:latest
+
 ```
 
 ## Basic Example

docs/pages/development/full-install.mdx

@@ -23,7 +23,7 @@ To install poetry, visit the [official Poetry website](https://python-poetry.org
   ```bash
   # See pyproject.toml for available extras
   # use "all" to include every optional dependency
-  poetry install -E parsing -E eval
+  poetry install -E eval
   ```
 
 ##### Configure Environment Variables:

docs/pages/getting-started/basic-example.mdx

@@ -5,7 +5,7 @@ This example demonstrates setting up a local server using the R2R framework to h
 
 ```bash
 # use the `'r2r[all]'` to download all required deps
-pip install 'r2r[parsing,eval]'
+pip install 'r2r[eval]'
 
 export OPENAI_API_KEY="replace with your openai key"
 export LOCAL_DB_PATH=local.sqlite

docs/pages/getting-started/quick-install.mdx

@@ -4,7 +4,7 @@ Install R2R swiftly using `pip` to get started with minimal setup. This method w
 
 ```bash
 # `'r2r[all]'` to download all optional deps
-pip install 'r2r[parsing,eval]'
+pip install 'r2r[eval]'
 
 # setup env 
 export OPENAI_API_KEY=sk-...

docs/pages/tutorials/local_rag.mdx

@@ -18,7 +18,7 @@ Next, let's install R2R itself. We'll use pip to manage our Python dependencies.
 
 ```bash
 pip install --upgrade pip
-pip install 'r2r[eval,parsing,local_llm]'
+pip install 'r2r[eval,local_llm]'
 ```
 
 This will install R2R along with the dependencies needed to run local LLMs.

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "r2r"
-version = "0.1.31"
+version = "0.1.32"
 description = "SciPhi R2R"
 authors = ["Owen Colegrove <owen@sciphi.ai>"]
 license = "MIT"
@@ -16,12 +16,14 @@ include = ["config.json"]
 python = ">=3.9,<3.13"
 
 # Required dependencies
+bs4= "^0.0.2"
 fastapi = "^0.109.2"
 fire = "^0.5.0"
 gunicorn = "^21.2.0"
 litellm = "^1.34.0"
 openai = "^1.11.1"
 pydantic = "^2.6.3"
+pypdf = "^4.2.0"
 python-dotenv = "^1.0.1"
 python-multipart = "^0.0.9"
 requests = "^2.31.0"
@@ -31,8 +33,6 @@ vecs = "^0.4.0"
 
 
 # Optional dependencies
-bs4 = {version = "^0.0.2", optional = true}
-pypdf = {version = "^4.0.2", optional = true}
 tiktoken = {version = "^0.5.2", optional = true}
 datasets = {version = "^2.16.1", optional = true}
 qdrant_client = {version = "^1.7.0", optional = true}
@@ -48,7 +48,6 @@ llama-cpp-python = {version = "^0.2.57", optional = true}
 sentence-transformers = {version = "^2.6.1", optional = true}
 
 [tool.poetry.extras]
-parsing = ["bs4", "pypdf"]
 embedding = ["tiktoken"]
 streaming = ["datasets"]
 qdrant = ["qdrant_client"]
@@ -59,6 +58,7 @@ eval = ["parea-ai"]
 ionic = ["ionic-api-sdk"]
 reducto = ["boto3"]
 exa = ["exa-py"]
+sentence_transformers = ["sentence-transformers"]
 local_llm = ["llama-cpp-python", "sentence-transformers"]
 all = ["bs4", "pypdf", "tiktoken", "datasets", "qdrant_client", "psycopg2-binary", "sentry-sdk", "parea-ai", "boto3", "exa-py", "llama-cpp-python"]
 

r2r/core/adapters/base.py

@@ -87,5 +87,5 @@ def adapt(self, data: bytes) -> list[str]:
                 page_text = "".join(
                     filter(lambda x: x in string.printable, page_text)
                 )
-                text += page_text
+                text += page_text + " "
         return [text]

r2r/eval/parea/base.py

@@ -37,7 +37,7 @@ def create_log(query: str, context: str, completion: str) -> Log:
             self._create_log = create_log
         except ImportError:
             raise ImportError(
-                "Parea is not installed. Please install it using `pip install parea`."
+                "Parea is not installed. Please install it using `pip install parea-ai`."
             )
         if not os.getenv("OPENAI_API_KEY"):
             raise ValueError(

r2r/examples/servers/basic_pipeline.py

@@ -14,6 +14,18 @@
     "local_llama_cpp": os.path.join(configs_path, "local_llama_cpp.json"),
 }
 
+
+def create_app(config_name: str = "default"):
+    config_path = OPTIONS[config_name]
+
+    app = E2EPipelineFactory.create_pipeline(
+        config=R2RConfig.load_config(config_path)
+    )
+    return app
+
+
+app = create_app()
+
 if __name__ == "__main__":
     parser = argparse.ArgumentParser(description="R2R Pipeline")
     parser.add_argument(
@@ -23,17 +35,6 @@
         choices=OPTIONS.keys(),
         help="Configuration option for the pipeline",
     )
-    args = parser.parse_args()
-
-    config_path = OPTIONS[args.config]
-
-    # Creates a pipeline with the specified configuration
-    # This is the main entry point for the application
-    # The pipeline is built using the specified configuration file
-    # Read more about the configuration in the documentation [https://r2r-docs.sciphi.ai/core-features/factory]
-    app = E2EPipelineFactory.create_pipeline(
-        config=R2RConfig.load_config(config_path)
-    )
+    args, _ = parser.parse_known_args()
 
-    # Run the FastAPI application using Uvicorn
     uvicorn.run(app, host="0.0.0.0", port=8000)

r2r/main/app.py

@@ -227,9 +227,10 @@ async def rag_completion(
                     "run_id": str(rag_pipeline.pipeline_run_info["run_id"]),
                     "settings": query.settings.rag_settings.dict(),
                 }
-                background_tasks.add_task(
-                    requests.post, f"{url}/eval", json=payload
-                )
+                if config.evals.get("frequency", 0.0) > 0.0:
+                    background_tasks.add_task(
+                        requests.post, f"{url}/eval", json=payload
+                    )
 
                 return rag_completion
 

r2r/pipelines/basic/eval.py

@@ -42,7 +42,7 @@ def __init__(
                 from r2r.eval import PareaEvalProvider
             except ImportError:
                 raise ImportError(
-                    "Parea is not installed. Please install it using `pip install parea`."
+                    "Parea is not installed. Please install it using `pip install parea-ai`."
                 )
             self.eval_provider = PareaEvalProvider(
                 provider,