fix: make environment variables more configurable

env.sample → .env.example

@@ -1,4 +1,3 @@
-CLIENT_URL="https://api.aleph-alpha.com"
 ARGILLA_API_URL="http://localhost:6900/"
 ARGILLA_API_KEY="argilla.apikey"
 
@@ -13,7 +12,14 @@ POSTGRES_DB=il_sdk
 POSTGRES_USER=il_sdk
 POSTGRES_PASSWORD=test
 
-# things to adapt
+# ---- Things to adapt ----
+CLIENT_URL=...
+AA_TOKEN=token
+DOCUMENT_INDEX_URL=...
+
+# needed for studio integration
 DATA_SERVICE_URL=...
+AUTHORIZATION_SERVICE_URL=...
+
+# needed for hugging face integration
 HUGGING_FACE_TOKEN=token
-AA_TOKEN=token

.github/workflows/sdk-tests.yml

@@ -147,9 +147,9 @@ jobs:
           POSTGRES_DB: "il_sdk"
           POSTGRES_USER: "il_sdk"
           POSTGRES_PASSWORD: "test"
-          AUTHORIZATION_SERVICE_URL: "none"
+          AUTHORIZATION_SERVICE_URL: ${{ secrets.AUTHORIZATION_SERVICE_URL }}
           AA_TOKEN: ${{ secrets.AA_TOKEN }}
-          API_SCHEDULER_URL: "https://api.aleph-alpha.com"
+          API_SCHEDULER_URL: ${{ secrets.CLIENT_URL }}
           DATA_SERVICE_URL: ${{secrets.DATA_SERVICE_URL}}
         credentials:
           username: "unused"
@@ -190,6 +190,7 @@ jobs:
           ARGILLA_API_KEY: "argilla.apikey"
           CLIENT_URL: ${{ secrets.CLIENT_URL }}
           STUDIO_URL: "http://localhost:8000/"
+          DOCUMENT_INDEX_URL: ${{secrets.DOCUMENT_INDEX_URL}}
           POSTGRES_HOST: "localhost"
           POSTGRES_PORT: "5433"
           POSTGRES_DB: "il_sdk"
@@ -235,9 +236,9 @@ jobs:
           POSTGRES_DB: "il_sdk"
           POSTGRES_USER: "il_sdk"
           POSTGRES_PASSWORD: "test"
-          AUTHORIZATION_SERVICE_URL: "none"
+          AUTHORIZATION_SERVICE_URL: ${{ secrets.AUTHORIZATION_SERVICE_URL }}
           AA_TOKEN: ${{ secrets.AA_TOKEN }}
-          API_SCHEDULER_URL: "https://api.aleph-alpha.com"
+          API_SCHEDULER_URL: ${{ secrets.CLIENT_URL }}
           DATA_SERVICE_URL: ${{secrets.DATA_SERVICE_URL}}
         credentials:
           username: "unused"
@@ -274,5 +275,6 @@ jobs:
           ARGILLA_API_KEY: "argilla.apikey"
           CLIENT_URL: ${{ secrets.CLIENT_URL }}
           STUDIO_URL: "http://localhost:8001"
+          DOCUMENT_INDEX_URL: ${{secrets.DOCUMENT_INDEX_URL}}
         run: |
           ./scripts/notebook_runner.sh

CHANGELOG.md

@@ -19,6 +19,9 @@
 
 ### Breaking Changes
  - The env variable `POSTGRES_HOST` is split into `POSTGRES_HOST` and `POSTGRES_PORT`. This affects all classes interacting with Studio and the `InstructionFinetuningDataRepository`.
+ - The following env variables now need to be set (previously pointed to defaults)
+   - `CLIENT_URL` - URL of your inference stack
+   - `DOCUMENT_INDEX_URL` - URL of the document index
 
 ## 8.0.0
 

README.md

@@ -116,7 +116,7 @@ The tutorials aim to guide you through implementing several common use-cases wit
 
 ### Setup LLM access
 
-The tutorials require access to an LLM endpoint. You can choose between using the Aleph Alpha API (`https://api.aleph-alpha.com`) or an on-premise setup by configuring the appropriate environment variables. To configure the environment variables, create a `.env` file in the root directory of the project and copy the contents of the `.env.sample` file into it.
+The tutorials require access to an LLM endpoint. You can choose between using the Aleph Alpha API (`https://api.aleph-alpha.com`) or an on-premise setup by configuring the appropriate environment variables. To configure the environment variables, create a `.env` file in the root directory of the project and copy the contents of the `.env.example` file into it.
 
 To use the **Aleph Alpha API**, that is set as the default host URL, set the `AA_TOKEN` variable to your [Aleph Alpha access token,](https://docs.aleph-alpha.com/docs/account/#create-a-new-token) and you are good to go.
 

docker-compose.yaml

@@ -71,8 +71,7 @@ services:
     env_file: ".env" # mainly for AA-TOKEN, DB User/PW
     environment:
       POSTGRES_HOST: postgres
-      AUTHORIZATION_SERVICE_URL: "none"
-      API_SCHEDULER_URL: "https://api.aleph-alpha.com"
+      API_SCHEDULER_URL: ${CLIENT_URL}
   postgres:
     image: postgres:15
     ports:

src/documentation/document_index.ipynb

@@ -25,7 +25,7 @@
     "    LimitedConcurrencyClient,\n",
     "    SemanticEmbed,\n",
     ")\n",
-    "from intelligence_layer.core import InMemoryTracer\n",
+    "from intelligence_layer.core import InMemoryTracer, LuminousControlModel\n",
     "from intelligence_layer.examples import MultipleChunkRetrieverQa, RetrieverBasedQaInput\n",
     "\n",
     "load_dotenv()"
@@ -61,9 +61,7 @@
    "source": [
     "## Upload documents to the Document Index\n",
     "\n",
-    "To search through the DI, you'll first need to upload the documents to it.\n",
-    "For now, we'll use the [DI instance hosted by Aleph Alpha](https://app.document-index.aleph-alpha.com).\n",
-    "We assume you have an assigned namespace and possess a token to access it."
+    "To search through the DI, you'll first need to upload the documents to it. We assume that the URL of your DI instance is available under the `DOCUMENT_INDEX_URL` environment variable, and that you already have a namespace and a token to access it."
    ]
   },
   {
@@ -72,8 +70,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "# specify this for your own namespace\n",
-    "NAMESPACE = \"aleph-alpha\""
+    "# change this to your namespace\n",
+    "NAMESPACE = \"Search\""
    ]
   },
   {
@@ -84,7 +82,7 @@
    "source": [
     "document_index = DocumentIndexClient(\n",
     "    token=getenv(\"AA_TOKEN\"),\n",
-    "    base_document_index_url=\"https://document-index.aleph-alpha.com\",\n",
+    "    base_document_index_url=getenv(\"DOCUMENT_INDEX_URL\"),\n",
     ")"
    ]
   },
@@ -630,7 +628,9 @@
    "outputs": [],
    "source": [
     "client = LimitedConcurrencyClient.from_env()\n",
-    "retriever_qa = MultipleChunkRetrieverQa(document_index_retriever, insert_chunk_number=3)\n",
+    "retriever_qa = MultipleChunkRetrieverQa(\n",
+    "    document_index_retriever, insert_chunk_number=3, model=LuminousControlModel()\n",
+    ")\n",
     "\n",
     "\n",
     "input = RetrieverBasedQaInput(\n",
@@ -659,18 +659,11 @@
    "source": [
     "tracer"
    ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": []
   }
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "Python 3 (ipykernel)",
+   "display_name": "intelligence-layer-LP3DLT23-py3.12",
    "language": "python",
    "name": "python3"
   },
@@ -684,7 +677,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.11.9"
+   "version": "3.12.2"
   }
  },
  "nbformat": 4,

src/documentation/elo_qa_eval.ipynb

@@ -27,9 +27,6 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "from os import getenv\n",
-    "\n",
-    "from aleph_alpha_client import Client\n",
     "from dotenv import load_dotenv\n",
     "\n",
     "from intelligence_layer.connectors import LimitedConcurrencyClient\n",
@@ -56,8 +53,7 @@
     "\n",
     "load_dotenv()\n",
     "\n",
-    "aa_client = Client(getenv(\"AA_TOKEN\"))\n",
-    "limited_concurrency_client = LimitedConcurrencyClient(aa_client, max_retry_time=60)"
+    "aa_client = limited_concurrency_client = LimitedConcurrencyClient.from_env()"
    ]
   },
   {
@@ -205,7 +201,7 @@
    "source": [
     "models = [\n",
     "    LuminousControlModel(name=\"luminous-base-control\", client=aa_client),\n",
-    "    LuminousControlModel(name=\"luminous-supreme-control\", client=aa_client),\n",
+    "    Llama3InstructModel(name=\"llama-3.1-8b-instruct\", client=aa_client),\n",
     "]\n",
     "\n",
     "for model in models:\n",
@@ -292,6 +288,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
+    "# Here we evaluate with the same model as we want to evaluate for the evaluation.\n",
+    "# This includes a significant bias and is generally less recommended.\n",
     "elo_qa_evaluation_logic = EloQaEvaluationLogic(\n",
     "    model=Llama3InstructModel(name=\"llama-3.1-8b-instruct\")\n",
     ")\n",
@@ -450,8 +448,7 @@
    "outputs": [],
    "source": [
     "newly_added_models = [\n",
-    "    LuminousControlModel(name=\"luminous-base-control-20230501\", client=aa_client),\n",
-    "    LuminousControlModel(name=\"luminous-supreme-control-20230501\", client=aa_client),\n",
+    "    Llama3InstructModel(name=\"llama-3.1-70b-instruct\", client=aa_client),\n",
     "]\n",
     "\n",
     "for model in newly_added_models:\n",

src/documentation/evaluate_with_studio.ipynb

@@ -84,13 +84,6 @@
     "Therefore, let's check out what it looks like."
    ]
   },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -126,14 +119,17 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "all_labels = list(set(item[\"label\"] for item in data))\n",
+    "# we grab only a subset of the data here to speed up the evaluation. Remove the index to run on all example datapoints.\n",
+    "subset_of_data = data[:5]\n",
+    "\n",
+    "all_labels = list(set(item[\"label\"] for item in subset_of_data))\n",
     "dataset = studio_dataset_repository.create_dataset(\n",
     "    examples=[\n",
     "        Example(\n",
     "            input=ClassifyInput(chunk=TextChunk(item[\"message\"]), labels=all_labels),\n",
     "            expected_output=item[\"label\"],\n",
     "        )\n",
-    "        for item in data\n",
+    "        for item in subset_of_data\n",
     "    ],\n",
     "    dataset_name=\"Single Label Classify Dataset\",\n",
     ")\n",
@@ -281,7 +277,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "intelligence-layer-ZqHLMTHE-py3.12",
+   "display_name": "intelligence-layer-LP3DLT23-py3.12",
    "language": "python",
    "name": "python3"
   },
@@ -295,7 +291,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.12.3"
+   "version": "3.12.2"
   }
  },
  "nbformat": 4,

src/documentation/fastapi_example.py

@@ -65,7 +65,7 @@ def __call__(
 def client() -> Client:
     return Client(
         token=os.environ["AA_TOKEN"],
-        host=os.getenv("AA_CLIENT_BASE_URL", "https://api.aleph-alpha.com"),
+        host=os.environ["CLIENT_URL"],
     )
 
 
@@ -78,7 +78,7 @@ def default_model(
 def summary_task(
     model: Annotated[LuminousControlModel, Depends(default_model)],
 ) -> SteerableSingleChunkSummarize:
-    return SteerableSingleChunkSummarize(model)
+    return SteerableSingleChunkSummarize(model=model)
 
 
 @app.post(

src/documentation/how_tos/example_data.py

@@ -112,6 +112,7 @@ class ExampleData:
     run_overview_2: RunOverview
     evaluation_overview_1: EvaluationOverview
     evaluation_overview_2: EvaluationOverview
+    studio_project_name: str
 
 
 def example_data() -> ExampleData:
@@ -159,6 +160,7 @@ def example_data() -> ExampleData:
     example_data.run_overview_2 = run_overview_2
     example_data.evaluation_overview_1 = evaluation_overview_1
     example_data.evaluation_overview_2 = evaluation_overview_2
+    example_data.studio_project_name = "My Example Project"
 
     return example_data
 

src/documentation/how_tos/how_to_aggregate_evaluations.ipynb

@@ -70,7 +70,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "intelligence-layer-aL2cXmJM-py3.11",
+   "display_name": "intelligence-layer-LP3DLT23-py3.12",
    "language": "python",
    "name": "python3"
   },
@@ -84,7 +84,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.11.8"
+   "version": "3.12.2"
   }
  },
  "nbformat": 4,