Text Generation
PEFT
Safetensors
English
lora
olmo
uncertainty
calibration
abstention
retrieval-augmented-generation
Instructions to use Sentovara/Sentinel-S1-32B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use Sentovara/Sentinel-S1-32B with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("allenai/Olmo-3.1-32B-Instruct") model = PeftModel.from_pretrained(base_model, "Sentovara/Sentinel-S1-32B") - Notebooks
- Google Colab
- Kaggle
Model card: note on the identity answer
Browse files
README.md
CHANGED
|
@@ -88,6 +88,7 @@ Our search took the top 50 BM25 passages from each of two Wikipedia indexes (the
|
|
| 88 |
- **Maths flagging is weak.** Sentinel flags only a small share of its own maths mistakes. Most maths errors are still stated plainly.
|
| 89 |
- **The marking is strict.** Answers are marked right only if they contain the expected answer, so some correct answers phrased differently are marked wrong.
|
| 90 |
- **Tested on public English question sets and everyday chat only.** It has not been tested on other kinds of work or other languages. Do not rely on it for medical, legal, financial or safety-critical decisions.
|
|
|
|
| 91 |
- **One model family.** The recipe works on the 7B and 32B Olmo models. It has not been tested on other model families.
|
| 92 |
|
| 93 |
## Credits and licences
|
|
|
|
| 88 |
- **Maths flagging is weak.** Sentinel flags only a small share of its own maths mistakes. Most maths errors are still stated plainly.
|
| 89 |
- **The marking is strict.** Answers are marked right only if they contain the expected answer, so some correct answers phrased differently are marked wrong.
|
| 90 |
- **Tested on public English question sets and everyday chat only.** It has not been tested on other kinds of work or other languages. Do not rely on it for medical, legal, financial or safety-critical decisions.
|
| 91 |
+
- **Identity answer.** If asked what model it is based on, the 32B may say Olmo 3 7B Instruct: its short identity training set was carried over from the 7B unchanged. It is built on Olmo 3.1 32B Instruct. This does not affect any of the results above.
|
| 92 |
- **One model family.** The recipe works on the 7B and 32B Olmo models. It has not been tested on other model families.
|
| 93 |
|
| 94 |
## Credits and licences
|