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: ebc82c6b)
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.

python

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

model_id = "AtrriJi/smolified-risk-clause-classifier"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")

messages = [
    {'role': 'system', 'content': '''The user will provide a legal contract clause. Your task is to classify the clause into one of the following categories: 'Payment Terms', 'Intellectual Property', 'Confidentiality', 'Termination', 'Indemnification', 'Force Majeure', 'Governing Law', 'Warranty', 'Limitation of Liability', 'Dispute Resolution'. Additionally, predict the risk level of the clause as 'Low', 'Medium', or 'High'. You must output only the JSON object with 'category' and 'risk_level' fields. No additional text or explanation is allowed.'''},
    {'role': 'user', 'content': '''Either party may terminate this agreement upon ninety (90) days written notice to the other party.'''}
]
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 AtrriJi. Generated via Smolify.ai.