GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
andreasmartin
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
andreasmartin
Model Tree
andreasmartin/apertus-v1.5-8b-text
Input Modalities
Output Modalities
Supported Functionality
Prompt:
text
You are an expert biomedical information extraction system. Extract all mentioned drugs and adverse drug events (ADEs) from the clinical narrative as JSON. Clinical Narrative:"A 64-year-old female patient with chronic atrial fibrillation presented with severe bradycardia, visual yellow halos, and nausea after increasing her digoxin dosage to 0.25 mg daily. Electrocardiogram confirmed digitalis toxicity. Concurrently, she had been prescribed amiodarone two weeks earlier."Model Output:
json
{ "drugs": ["digoxin", "amiodarone"], "adverse_drug_events": ["severe bradycardia", "visual yellow halos", "nausea"]}Prompt:
text
You are an expert biomedical information extraction system. Extract all mentioned drugs and adverse drug events (ADEs) from the clinical narrative as JSON. Clinical Narrative:"The patient was initiated on metformin 500 mg twice daily for type 2 diabetes mellitus and reported good glycemic control without any adverse symptoms."Model Output:
json
{ "drugs": ["metformin"], "adverse_drug_events": []}bash
vllm serve andreasmartin/apertus-1.5-8b-biomedical-ade-sft-exercise-60steps-20260925-193431 --tensor-parallel-size 1 --gpu-memory-utilization 0.9python
from transformers import AutoModelForCausalLM, AutoTokenizer tokenizer = AutoTokenizer.from_pretrained("andreasmartin/apertus-1.5-8b-biomedical-ade-sft-exercise-60steps-20260925-193431")model = AutoModelForCausalLM.from_pretrained("andreasmartin/apertus-1.5-8b-biomedical-ade-sft-exercise-60steps-20260925-193431", device_map="auto")| Parameter | Value |
|---|---|
| Run ID | 20260925-193431 |
| Run Label | exercise |
| Base Model | andreasmartin/apertus-v1.5-8b-text |
| Dataset | mihirhirave/entity_extraction_ade_v2_with_validation |
| Dataset Slice | train[:600] (Train) / validation[:100] (Validation) |
| Optimizer Steps | 60 |
| Effective Batch Size | 8 (Per-device batch size 2 x Gradient Accumulation 2 x 2 GPUs) |
| Learning Rate | 0.0002 (Cosine decay schedule) |
| Merge Method | merged_16bit (Standalone 16-bit Safetensors) |
| Final Training Loss | 0.0976 |
| Final Validation Loss | 0.0612 |
| Best Validation Loss | 0.0609 |
| W&B Tracking | fhnw-ch/Apertus / apertus-1.5-8b-biomedical-ade-sft-exercise-60steps-20260925-193431 |
You are an expert biomedical information extraction system. Extract all mentioned drugs and adverse drug events (ADEs) from the clinical narrative as JSON.Clinical Narrative:"A 64-year-old female patient with chronic atrial fibrillation presented with severe bradycardia, visual yellow halos, and nausea after increasing her digoxin dosage to 0.25 mg daily. Electrocardiogram confirmed digitalis toxicity. Concurrently, she had been prescribed amiodarone two weeks earlier."{ "drugs": ["digoxin", "amiodarone"], "adverse_drug_events": ["severe bradycardia", "visual yellow halos", "nausea"]}You are an expert biomedical information extraction system. Extract all mentioned drugs and adverse drug events (ADEs) from the clinical narrative as JSON.Clinical Narrative:"The patient was initiated on metformin 500 mg twice daily for type 2 diabetes mellitus and reported good glycemic control without any adverse symptoms."{ "drugs": ["metformin"], "adverse_drug_events": []}vllm serve andreasmartin/apertus-1.5-8b-biomedical-ade-sft-exercise-60steps-20260925-193431 --tensor-parallel-size 1 --gpu-memory-utilization 0.9from transformers import AutoModelForCausalLM, AutoTokenizer tokenizer = AutoTokenizer.from_pretrained("andreasmartin/apertus-1.5-8b-biomedical-ade-sft-exercise-60steps-20260925-193431")model = AutoModelForCausalLM.from_pretrained("andreasmartin/apertus-1.5-8b-biomedical-ade-sft-exercise-60steps-20260925-193431", device_map="auto")