🤏 smolified-clinical-scribe

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: e66b366f)
• 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-clinical-scribe"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")

messages = [
    {'role': 'system', 'content': '''The user will provide a medical transcript. Your task is to extract information and structure it into a SOAP note format. Only include information explicitly stated in the transcript. Do not add any new medical information or make assumptions.'''},
    {'role': 'user', 'content': '''Doctor: Hello, Mr. Smith. What's troubling you? Patient: Been feeling pretty tired, doctor, for about a month now. And a bit down. I'm just dragging all the time. Can't focus at work either. Not sleeping great, keep waking up early. Doctor: Any other changes? Appetite, weight? Patient: Yeah, lost about 5 pounds without trying. And my appetite is lower. No motivation, you know? Doctor: Okay. *examines patient* Heart sounds normal, lungs clear. No significant findings on abdominal exam. Looks generally well. Mood appears somewhat flat. Doctor: Your symptoms sound consistent with depression, and perhaps some underlying fatigue. Let's start you on an antidepressant, a low dose of escitalopram, and I'd like you to consider some talk therapy. We'll check in in two weeks to see how you're doing. Try to maintain a regular sleep schedule, even if you're not sleeping through the night. Avoid caffeine late in the day.'''}
]
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.