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: 22fdd899)
Architecture: gemma-3-270m
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, and Hugging Face Transformers.

python

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

model_id = "Shubhankar444/smolified-ingredient-extractor"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")

messages = [
    {"role": "system", "content": '''Extract ingredients from the Indian recipe and return as a Python list.'''},
    {"role": "user", "content": '''Prepare Bhindi Masala by slicing okra and shallow frying in vegetable oil. Keep aside. In the same pan, saute onions and tomatoes with cumin, coriander powder, dry mango powder, and salt. Add the okra back in and toss well until the vegetables are soft. Serve with hot roti.'''}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize = False,
    add_generation_prompt = True,
)
if "gemma-3-270m" == "gemma-3-270m":
    text = text.removeprefix('<bos>')

from transformers import TextStreamer
_ = model.generate(
    **tokenizer(text, return_tensors = "pt").to(model.device),
    max_new_tokens = 1000,
    temperature = 1.0, top_p = 0.95, top_k = 64,
    streamer = TextStreamer(tokenizer, skip_prompt = True),
)

⚖️ License & Ownership

This model weights are a sovereign asset owned by Shubhankar444. Generated via Smolify.ai.