Upgrade

README.md

@@ -4,7 +4,7 @@ This project provides all the code necessary to deploy the APE classification mo
 
 ## Prerequisites
 
-- Python 3.12
+- Python 3.13
 
 ## Setup
 

pyproject.toml

@@ -5,21 +5,21 @@ description = "API for deployment of the classification model for APE nomenclatu
 readme = "README.md"
 requires-python = ">=3.13"
 dependencies = [
-    "fastapi>=0.115.12",
-    "mlflow>=2.21.2",
-    "nltk>=3.9.1",
-    "numpy>=2.2.4",
+    "fastapi>=0.138.1",
+    "mlflow>=3.14.0",
+    "nltk>=3.9.4",
+    "numpy>=2.5.0",
     "omegaconf>=2.3.0",
-    "pandas>=2.2.3",
-    "pendulum>=3.0.0",
-    "pyarrow>=19.0.1",
-    "pydantic>=2.11.1",
-    "requests>=2.32.3",
-    "s3fs>=2025.3.2",
-    "torch>=2.6.0",
+    "pandas>=2.3.3",
+    "pendulum>=3.2.0",
+    "pyarrow>=24.0.0",
+    "pydantic>=2.13.4",
+    "requests>=2.34.2",
+    "s3fs>=2026.4.0",
+    "torch>=2.12.1",
     "torchfasttext",
-    "tqdm>=4.67.1",
-    "unidecode>=1.3.8",
+    "tqdm>=4.68.3",
+    "unidecode>=1.4.0",
     "uvicorn>=0.34.0",
 ]
 authors = [

setup.sh

@@ -7,7 +7,7 @@ uv run pre-commit install
 export MLFLOW_S3_ENDPOINT_URL="https://$AWS_S3_ENDPOINT"
 export MLFLOW_TRACKING_URI=https://projet-ape-mlflow.user.lab.sspcloud.fr
 export MLFLOW_MODEL_NAME=FastText-pytorch-2025
-export MLFLOW_MODEL_VERSION="6"
+export MLFLOW_MODEL_VERSION="12"
 export API_USERNAME=username
 export API_PASSWORD=password
 export AUTH_API=False

src/api/main.py

@@ -24,7 +24,7 @@ async def lifespan(app: FastAPI):
     logger.info("🚀 Starting API lifespan")
 
     app.state.model = load_model()
-    app.state.run_id = app.state.model.metadata.run_id
+    app.state.model_id = app.state.model.metadata.model_id
 
     yield
     logger.info("🛑 Shutting down API lifespan")

src/api/models/responses.py

@@ -13,7 +13,7 @@ class OutputResponse(RootModel[Dict[str, Union[Prediction, float, str]]]):
     """
     Contract for the output response of the API including:
     - KV of PredictionResponse: normalized prediction responses generated by the model artifact
-    - MLversion: run_id as version of the ML model
+    - MLversion: model_id as version of the ML model
 
     Expected flat structure after normalization:
 
@@ -22,7 +22,7 @@ class OutputResponse(RootModel[Dict[str, Union[Prediction, float, str]]]):
     "2": Prediction,
     ...,
     "IC": float,      # required confidence score
-    "MLversion": str  # required run_id as model version
+    "MLversion": str  # required model_id as model version
     }
 
     Notes:

src/api/routes/predict.py

@@ -57,6 +57,6 @@ async def predict(
 
     output = request.app.state.model.predict(input_data, params=params_dict)
     return [
-        OutputResponse({**out.model_dump(), "MLversion": request.app.state.run_id})
+        OutputResponse({**out.model_dump(), "MLversion": request.app.state.model_id})
         for out in output
     ]

src/utils/load_model.py

@@ -5,10 +5,11 @@
 
 
 def load_model():
+    # mlflow.set_tracking_uri(os.environ["MLFLOW_TRACKING_URI"])
     model_uri = f"models:/{os.environ['MLFLOW_MODEL_NAME']}/{os.environ['MLFLOW_MODEL_VERSION']}"
 
     # Step 1: Set the destination path for the model artifacts
-    dst_path = "/tmp/my_model"
+    dst_path = "/tmp/my_model/artifacts/pyfunc_model"
 
     # Step 2: Download/extract the model here *without loading it yet*
     mlflow.artifacts.download_artifacts(artifact_uri=model_uri, dst_path=dst_path)

src/utils/logging.py

@@ -11,6 +11,7 @@ def configure_logging():
             logging.StreamHandler(),
         ],
     )
+    logging.getLogger("mlflow").setLevel(logging.ERROR)
 
 
 def log_prediction(query: dict, response: OutputResponse, index: int = 0):