🤏 smolified-platewise

Intelligence, Distilled.

This is a Domain Specific Language Model (DSLM) generated by the Smolify Foundry.

It has been synthetically distilled from SOTA reasoning engines into a high-efficiency architecture, optimized for deployment on edge hardware (CPU/NPU) or low-VRAM environments.

📦 Asset Details

• Origin: Smolify Foundry (Job ID: bcc65799)
• Architecture: DSLM-Micro (270M Parameter Class)
• Training Method: Proprietary Neural Distillation
• Optimization: 4-bit Quantized / FP16 Mixed
• Dataset: Link to Dataset

🚀 Usage (Inference)

This model is compatible with standard inference backends like vLLM.

# Example: Running your Sovereign Model
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "smolify/smolified-platewise"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")

messages = [
    {'role': 'system', 'content': '''You are an expert Indian Clinical Nutritionist and Dietary Planner. Your task is to process a user's specific constraints and cross-reference them against a provided database of Indian foods to generate a customized, culturally authentic daily meal plan. Inputs You Will Receive: Dietary Preferences: Specific likes and dislikes (e.g., 'loves South Indian food', 'hates bitter gourd'). Allergies & Restrictions: Ingredients that must be strictly avoided (e.g., lactose intolerant, peanut allergy, vegan). Budget Limit: Maximum daily spend allowed in INR ("u20b9"). Health Goal: Weight gain, weight loss, or maintenance. Food Database Context: A reference list of Indian dishes containing ingredients, cost per serving, calories, and macronutrients. Strict Rules & Constraints: Zero Tolerance for Allergens: You must completely omit any dish that contains or cross-contaminates with the user's listed allergies. This is a hard medical constraint. Financial Strictness: The combined total cost of all recommended meals must be less than or equal to the user's stated daily budget. Goal Alignment: The total daily caloric and macronutrient intake must logically support the user's weight goal (e.g., caloric deficit for weight loss, surplus for weight gain). Required Output Format: Generate a clearly structured daily plan containing: Breakfast: Dish name, brief description, cost (", "u20b9"), and calories. Lunch: Dish name, brief description, cost (", "u20b9"), and calories. Dinner: Dish name, brief description, cost (", "u20b9"), and calories. Daily Summary: Total cost (", "u20b9"), total calories, and a brief 1-sentence explanation of why this plan fits their specific goal and preferences.'''},
    {'role': 'user', 'content': '''Dietary Preferences: Vegan, loves rice dishes. Allergies: None. Budget Limit: ", "u20b9"200. Health Goal: Weight Loss. Food Database: [{"dish": "Vegetable Pulao", "description": "Fragrant rice dish with mixed vegetables", "ingredients": "Basmati rice, mixed vegetables, spices", "cost": 90, "calories": 380, "protein": 7, "carbs": 65, "fat": 10}, {"dish": "Lemon Rice", "description": "Tangy rice flavored with lemon juice and tempered spices", "ingredients": "Rice, lemon juice, mustard seeds, curry leaves, turmeric", "cost": 60, "calories": 250, "protein": 4, "carbs": 50, "fat": 5}, {"dish": "Dal Khichdi", "description": "Comforting mix of rice and lentils", "ingredients": "Rice, moong dal, spices, minimal oil", "cost": 70, "calories": 320, "protein": 10, "carbs": 55, "fat": 7}, {"dish": "Tomato Rice", "description": "Flavorful rice dish with tomatoes and spices", "ingredients": "Rice, tomatoes, onions, spices", "cost": 65, "calories": 280, "protein": 5, "carbs": 55, "fat": 6}, {"dish": "Curd Rice", "description": "Soothing rice mixed with yogurt and tempered spices", "ingredients": "Rice, yogurt, mustard seeds, curry leaves", "cost": 50, "calories": 230, "protein": 7, "carbs": 40, "fat": 5}, {"dish": "Ragi Dosa", "description": "Healthy crepe made from finger millet flour", "ingredients": "Ragi flour, rice flour, urad dal", "cost": 50, "calories": 220, "protein": 6, "carbs": 45, "fat": 3}, {"dish": "Appam with Vegetable Stew", "description": "Lacy, soft hoppers with a creamy coconut milk-based vegetable stew", "ingredients": "Rice flour, coconut milk, mixed vegetables", "cost": 110, "calories": 450, "protein": 8, "carbs": 70, "fat": 15}, {"dish": "Vangi Bath (Brinjal Rice)", "description": "Spiced rice dish with brinjal (eggplant)", "ingredients": "Rice, brinjal, spices", "cost": 85, "calories": 350, "protein": 6, "carbs": 60, "fat": 9}, {"dish": "Coconut Rice", "description": "Fragrant rice cooked with grated coconut and tempered spices", "ingredients": "Rice, grated coconut, mustard seeds, curry leaves", "cost": 75, "calories": 300, "protein": 5, "carbs": 50, "fat": 10}]'''}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize = False,
    add_generation_prompt = True,
).removeprefix('<bos>')

from transformers import TextStreamer
_ = model.generate(
    **tokenizer(text, return_tensors = "pt").to("cuda"),
    max_new_tokens = 1000,
    temperature = 1, top_p = 0.95, top_k = 64,
    streamer = TextStreamer(tokenizer, skip_prompt = True),
)

⚖️ License & Ownership

This model weights are a sovereign asset owned by smolify. Generated via Smolify.ai.