YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
MediScribe Medical Adapter
MediScribe Medical Adapter is a LoRA adapter for rural-clinic medical triage support. It is trained with Unsloth SFT on curated MediScribe clinical cases and is intended for clinical decision support demos, not autonomous diagnosis.
Base Model
- Base model:
unsloth/gemma-4-E4B-it-unsloth-bnb-4bit - Fine-tuning method: Unsloth LoRA SFT
- Adapter rank: 16
- Max sequence length: 2048
- Training rows: 26
- Evaluation rows: 6
Benchmark Snapshot
- Benchmark cases: 26
- Primary diagnosis hit rate: 0%
- Top-3 diagnosis hit rate: 0%
- Red-flag recall: 100%
- Average inference seconds: 0.00
Safety Position
The adapter should be used only with MediScribe's deterministic safety layer: red-flag rules, referral fallback, audit logs, and plain-language explanations. It does not replace a licensed clinician.
Naming And Attribution
This model artifact is named MediScribe Medical Adapter. It is based on a Gemma 4 model, but MediScribe is independent and is not endorsed by Google.
Gemma is a trademark of Google LLC.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support