Sema 1d endpoint

apps/fastapi/src/app/core/config.py

@@ -44,6 +44,10 @@ def assemble_cors_origins(cls, v: Union[str, List[str]]) -> Union[List[str], str
     SUPERUSER_PASSWORD: str = Field(..., env="SUPERUSER_PASSWORD")
     JWT_SECRET: str = Field(..., env="JWT_SECRET")
     S3_BUCKET_NAME: str = Field(default="fastapi-csv", env="S3_BUCKET_NAME")
+    SAGEMAKER_ENDPOINT_NAME: str = Field(
+        default="huggingface-pytorch-inference-2024-10-16-20-16-41-824",
+        env="SAGEMAKER_ENDPOINT_NAME",
+    )
 
     # Optional
     HUGGINGFACE_ACCESS_TOKEN: str = Field(None, env="HUGGINGFACE_ACCESS_TOKEN")

apps/fastapi/src/app/services/background_tasks.py

@@ -4,11 +4,13 @@
 from fastapi import HTTPException
 from supabase._async.client import AsyncClient
 
+from app.core.config import settings
 from app.crud.crud_conformational_b_prediction import crud_conformational_b_prediction
 from app.crud.crud_job import crud_job
 from app.crud.crud_linear_b_prediction import crud_linear_b_prediction
 from app.crud.crud_mhc_i_prediction import crud_mhc_i_prediction
 from app.crud.crud_mhc_ii_prediction import crud_mhc_ii_prediction
+from app.services.pipeline import run
 from app.services.postprocess import (
     process_conformational_b_prediction,
     process_linear_b_prediction,
@@ -35,6 +37,19 @@ async def process_and_update_prediction(
         # Update job status to 'running'
         await crud_job.update_status(db=db, id=job_id, status="running")
 
+        # Sample input
+        glp_1_receptor = "MAGAPGPLRLALLLLGMVGRAGPRPQGATVSLWETVQKWREYRRQCQRSLTEDPPPATDLFCNRTFDEYACWPDGEPGSFVNVSCPWYLPWASSVPQGHVYRFCTAEGLWLQKDNSSLPWRDLSECEESKRGERSSPEEQLLFLYIIYTVGYALSFSALVIASAILLGFRHLHCTRNYIHLNLFASFILRALSVFIKDAALKWMYSTAAQQHQWDGLLSYQDSLSCRLVFLLMQYCVAANYYWLLVEGVYLYTLLAFSVLSEQWIFRLYVSIGWGVPLLFVVPWGIVKYLYEDEGCWTRNSNMNYWLIIRLPILFAIGVNFLIFVRVICIVVSKLKANLMCKTDIKCRLAKSTLTLIPLLGTHEVIFAFVMDEHARGTLRFIKLFTELSFTSFQGLMVAILYCFVNNEVQLEFRKSWERWRLEHLHIQRDSSMKPLKCPTSSLSSGATAGSSMYTATCQASCS"
+        sample = {"inputs": glp_1_receptor}
+
+        # Run prediction pipeline
+        predictions = run(
+            input_objects=[sample],
+            endpoint_name=settings.SAGEMAKER_ENDPOINT_NAME,
+            model_type="single",
+        )
+
+        logger.info(predictions)
+
         # Process the prediction based on its type
         if prediction_type == "conformational-b":
             if is_structure_based:

apps/fastapi/src/app/services/inference.py

@@ -1,7 +1,6 @@
 import json
 
 import boto3
-import pandas as pd
 
 from app.core.config import settings
 
@@ -22,12 +21,11 @@ def get_predictions(requests, endpoint_name, model_name, model_type="mme"):
             )
             res = json.loads(response["Body"].read().decode("utf-8"))[0]
         else:
-            # XGBoost by default does not allow application/json
             response = sagemaker_runtime.invoke_endpoint(
-                ContentType="csv",
+                ContentType="application/json",
                 EndpointName=endpoint_name,
-                Body=pd.DataFrame(data=[request]).to_csv(index=False, header=False),
+                Body=json.dumps(request),
             )
-            res = float(response["Body"].read().decode("utf-8"))
+            res = json.loads(response["Body"].read().decode("utf-8"))
         responses.append(res)
     return responses

apps/fastapi/src/app/services/pipeline.py

@@ -1,14 +1,11 @@
-from typing import List
+from typing import Any, List
 
 from app.core.utils import get_endpoint
-from app.schemas.input_object import InputObject
 from app.services.inference import get_predictions
-from app.services.postprocess import do_postprocessing
-from app.services.preprocess import do_preprocess
 
 
 def run(
-    input_objects: List[InputObject],
+    input_objects: List[Any],
     endpoint_name: str,
     model_name=None,
     model_type="mme",
@@ -30,12 +27,18 @@ def run(
     if model_name is None:
         model_name = endpoint_name
 
-    preprocessed_df = do_preprocess(input_objects)
+    # Preprocess the inputs
+    preprocess = [obj for obj in input_objects]
+
+    # Run inference
     model_responses = get_predictions(
-        preprocessed_df.values.tolist(),
+        preprocess,
         endpoint_name,
         model_name,
         model_type=model_type,
     )
-    predictions = do_postprocessing(model_responses)
+
+    # Postprocess the inferences
+    predictions = model_responses
+
     return predictions