Upload 3 files

Recipe_maker/app.py

@@ -0,0 +1,149 @@
+import streamlit as st
+import pandas as pd
+import chromadb
+from sentence_transformers import SentenceTransformer
+import requests
+from PIL import Image
+from io import BytesIO
+from transformers import pipeline
+
+# --- 1. Setup Hugging Face API for Mistral ---
+huggingface_api_key = st.secrets["HUGGINGFACE_API_KEY"]# Replace with your Hugging Face API key
+model_name = "mistralai/Mistral-7B-Instruct-v0.1"
+
+def generate_mistral_response(prompt):
+    """Generates text using Mistral from Hugging Face API."""
+    api_url = f"https://api-inference.huggingface.co/models/{model_name}"
+    headers = {"Authorization": f"Bearer {huggingface_api_key}"}
+    payload = {"inputs": prompt}
+
+    response = requests.post(api_url, headers=headers, json=payload)
+    if response.status_code == 200:
+        return response.json()[0]["generated_text"]
+    else:
+        return f"Error: {response.json()}"
+
+# Initialize embedding model
+embedding_model = SentenceTransformer('all-mpnet-base-v2')
+
+# Initialize ChromaDB
+chroma_client = chromadb.PersistentClient(path="./chroma_db")
+collection = chroma_client.get_or_create_collection(name="recipe_collection")
+
+# --- 2. Load Dataset ---
+dataset_path = "recipes.csv"
+try:
+    recipes_df = pd.read_csv(dataset_path)
+except FileNotFoundError:
+    st.error(f"Error: File not found at {dataset_path}")
+    st.stop()
+
+# Fix column names (rename 'recipe_name' to 'title')
+if "recipe_name" in recipes_df.columns:
+    recipes_df.rename(columns={"recipe_name": "title"}, inplace=True)
+
+# Ensure dataset contains required columns
+required_columns = {"title", "ingredients", "img_src"}
+missing_columns = required_columns - set(recipes_df.columns)
+if missing_columns:
+    st.error(f"Missing columns in dataset: {missing_columns}")
+    st.stop()
+
+# Clean dataset
+recipes_df.fillna("", inplace=True)
+recipes_df["ingredients"] = recipes_df["ingredients"].str.lower().str.replace(r'[^\w\s]', '', regex=True)
+recipes_df["combined_text"] = recipes_df["title"] + " " + recipes_df["ingredients"]
+
+# --- 3. Generate Embeddings and Store in ChromaDB ---
+def get_sentence_transformer_embeddings(text):
+    return embedding_model.encode(text).tolist()
+
+# Ensure ChromaDB collection exists
+try:
+    existing_data = collection.get()
+    existing_ids = set(existing_data["ids"]) if existing_data and "ids" in existing_data else set()
+except Exception as e:
+    st.error(f"⚠️ Error fetching ChromaDB data: {e}")
+    existing_ids = set()
+
+# Add new embeddings only if not already in ChromaDB
+for index, row in recipes_df.iterrows():
+    recipe_id = str(index)
+    if recipe_id in existing_ids:
+        continue
+
+    embedding = get_sentence_transformer_embeddings(row["combined_text"])
+    if embedding:
+        collection.add(
+            embeddings=[embedding],
+            documents=[row["combined_text"]],
+            ids=[recipe_id]
+        )
+
+# --- 4. Retrieval Function ---
+def retrieve_recipes(query, top_k=3):
+    """Retrieves most relevant recipes."""
+    query_embedding = get_sentence_transformer_embeddings(query)
+    results = collection.query(query_embeddings=[query_embedding], n_results=top_k)
+
+    if results and "documents" in results and results["documents"]:
+        recipe_indices = [int(id) for id in results["ids"][0]]
+        return recipes_df.iloc[recipe_indices] if recipe_indices else None
+    return None
+
+# --- 5. Generate AI Recipe using Mistral ---
+def generate_recipe(user_query, retrieved_recipes):
+    """Generates a recipe using Mistral API."""
+    try:
+        relevant_ingredients = set()
+        for _, recipe in retrieved_recipes.iterrows():
+            relevant_ingredients.update(recipe["ingredients"].split())
+
+        structured_prompt = (
+            f"Create a new recipe using these ingredients: {', '.join(relevant_ingredients)}.\n"
+            f"Make sure the recipe matches: {user_query}.\n"
+            "- Recipe Name\n- Ingredients\n- Steps\n"
+        )
+
+        response = generate_mistral_response(structured_prompt)
+        return response if response else "No response from API."
+    except Exception as e:
+        return f"Error generating recipe: {e}"
+
+# --- 6. Display Image Function ---
+def display_image(image_url, recipe_name):
+    """Fetches and displays an image from a URL."""
+    try:
+        image_response = requests.get(image_url)
+        image_response.raise_for_status()
+        image = Image.open(BytesIO(image_response.content))
+        st.image(image, caption=recipe_name, use_container_width=True)
+    except requests.exceptions.RequestException:
+        st.warning(f"⚠️ Could not fetch image for {recipe_name}")
+
+# --- 7. Streamlit UI ---
+st.title("🍽️ AI Recipe Generator with Mistral")
+
+user_query = st.text_input("Enter a dish name or ingredients:", "")
+
+if st.button("Find Recipe"):
+    if user_query:
+        retrieved_recipes = retrieve_recipes(user_query)
+
+        if retrieved_recipes is not None and not retrieved_recipes.empty:
+            st.subheader("🍴 Found Recipes:")
+            for _, recipe in retrieved_recipes.iterrows():
+                st.markdown(f"### {recipe['title']}")
+                st.write(f"**Ingredients:** {recipe['ingredients']}")
+
+                if "img_src" in recipe and recipe["img_src"].strip():
+                    display_image(recipe["img_src"], recipe["title"])
+                else:
+                    st.warning("⚠️ No image available")
+
+            generated_recipe = generate_recipe(user_query, retrieved_recipes)
+
+            st.subheader("📝 AI-Generated Recipe:")
+            st.write(generated_recipe)
+        else:
+            st.warning("⚠️ No relevant recipes found.")

Recipe_maker/requirements.txt

@@ -0,0 +1,12 @@
+google-generativeai
+pandas
+numpy
+chromadb
+sentence-transformers
+streamlit 
+torch 
+transformers 
+diffusers
+accelerate 
+pillow 
+requests