Upload 5 files

dockerfile

@@ -0,0 +1,24 @@
+# Use the official Python image
+FROM python:3.9-slim
+
+# Set the working directory in the container
+WORKDIR /app
+
+# Install system dependencies
+RUN apt-get update && apt-get install -y \
+    git \
+    && apt-get clean
+
+# Install Python dependencies
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+
+# Copy the application code
+COPY . .
+
+# Expose the application port
+EXPOSE 8000
+
+# Run the application
+CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "8000"]

main.py

@@ -0,0 +1,29 @@
+# main.py
+
+from fastapi import FastAPI, File, UploadFile
+from transformers import pipeline
+from PIL import Image
+import io
+
+from prompts import generate_health_feedback_prompt
+from utils import send_prompt_to_llm
+
+app = FastAPI()
+
+pipe = pipeline("image-classification", model="nateraw/food")
+
+@app.post("/classify/")
+async def classify_image(file: UploadFile = File(...)):
+    # Read the uploaded image
+    image_bytes = await file.read()
+    image = Image.open(io.BytesIO(image_bytes))
+
+    # Classify the image to get ingredients
+    result = pipe(image)
+    ingredients = [res['label'] for res in result]
+
+    # Generate prompt and get feedback from the LLM
+    prompt = generate_health_feedback_prompt(ingredients)
+    health_feedback = send_prompt_to_llm(prompt)
+
+    return {"ingredients": ingredients, "health_feedback": health_feedback}

prompts.py

@@ -0,0 +1,21 @@
+# prompts.py
+
+json_structure = {
+    "meal": {
+        "ingredients": ["list of ingredients here"]
+    },
+    "response_format": {
+        "carbs": "numeric value in grams",
+        "fats": "numeric value in grams",
+        "proteins": "numeric value in grams",
+        "fibers": "numeric value in grams",
+        "feedback": "One-sentence feedback about the healthiness of the meal and suggestions for improvement"
+    }
+}
+
+def generate_health_feedback_prompt(ingredients):
+    return f"""
+    Given this list of ingredients: {ingredients}, estimate the macronutrient content and return only the values as a JSON object.
+    Specify the amount of carbs, protein, fat, and fiber in grams, and include a one-sentence feedback on the meal healthiness.
+    The response should strictly follow this JSON format: {json_structure}. Please respond with only the JSON object.
+    """

requirements.txt

@@ -0,0 +1,6 @@
+fastapi==0.74.*
+requests==2.27.*
+uvicorn[standard]==0.17.*
+sentencepiece==0.1.*
+torch==1.11.*
+transformers==4.*

utils.py

@@ -0,0 +1,14 @@
+# utils.py
+
+import requests
+
+def send_prompt_to_llm(prompt, model="llama3", endpoint="http://localhost:11434/v1/chat/completions"):
+    response = requests.post(
+        endpoint,
+        json={
+            "model": model,
+            "messages": [{"role": "user", "content": prompt}],
+        },
+    )
+    response.raise_for_status()  # Raise an error for failed requests
+    return response.json()["choices"][0]["message"]["content"]