Downloads · 30 days
15
29% of all-time downloads
gemma-3-4b-pt-SFT
medAdapt/gemma-3-4b-pt-SFT
gemma-3-4b-pt-SFT is a question answering model from medAdapt. Use it when the input is a question plus a passage. The card lists the license as apache-2.0.
This checkpoint is a supervised fine-tuned (SFT) version of gemma-3-4b-pt, trained on labeled medical QA (MedInjection-FR/ALL) data to improve task performance—especially for medical MCQA. SFT is implemented using DoRA.
All-time downloads
51
Public
Parameters
4.3B
17.2 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors17.2 GB · 100%
At a glance
- Task
- Question Answering
- License
- apache-2.0
- Model type
- gemma3
- Access
- Public
- Created
- Jan 4, 2026
- Updated
- Jan 4, 2026
- SHA
- 97e457c9
Try a prompt
Base models
Datasets
Model card
From the Hugging Face model README
gemma-3-4b-pt-SFT (SFT)
Model description
This checkpoint is a supervised fine-tuned (SFT) version of gemma-3-4b-pt, trained on labeled medical QA (MedInjection-FR/ALL) data to improve task performance—especially for medical MCQA. SFT is implemented using DoRA.
- Task
- Question Answering
- Type
- gemma3
- License
- apache-2.0
- Languages
- fr, en
- Created
- Jan 4, 2026
- Updated
- Jan 4, 2026