Instructions to use sassoftware/gemma-4-12B-it with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use sassoftware/gemma-4-12B-it with Transformers:
# Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("sassoftware/gemma-4-12B-it") model = AutoModelForMultimodalLM.from_pretrained("sassoftware/gemma-4-12B-it", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Added SAS Notes to Model Card
#1
by nadolsw - opened
README.md
CHANGED
|
@@ -524,4 +524,21 @@ The development of vision-language models (VLMs) raises several ethical concerns
|
|
| 524 |
|
| 525 |
### **Benefits**
|
| 526 |
|
| 527 |
-
At the time of release, this family of models provides high-performance open vision-language model implementations designed from the ground up for responsible AI development compared to similarly sized models.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 524 |
|
| 525 |
### **Benefits**
|
| 526 |
|
| 527 |
+
At the time of release, this family of models provides high-performance open vision-language model implementations designed from the ground up for responsible AI development compared to similarly sized models.
|
| 528 |
+
|
| 529 |
+
## Source Model Repo
|
| 530 |
+
|
| 531 |
+
This model is derived from [`google/gemma-4-12B-it`](https://huggingface.co/google/gemma-4-12B-it).
|
| 532 |
+
See the upstream repository for the original safetensors weights, training data, and the full upstream model card.
|
| 533 |
+
|
| 534 |
+
## SAS Modifications
|
| 535 |
+
|
| 536 |
+
SAS has not made any modifications to the source model weights or file artifacts.
|
| 537 |
+
|
| 538 |
+
## SAS Product Integration
|
| 539 |
+
|
| 540 |
+
This repo is consumed by the SAS Document Analysis product offering
|
| 541 |
+
|
| 542 |
+
## Model Security Scan
|
| 543 |
+
|
| 544 |
+
Model weights were scanned on 2026-07-08. All checks passed. Safetensor file recevied a LOW risk level assessment.
|