Text Classification
Transformers
ONNX
Safetensors
English
modernbert
rag
governance
hallucination-detection
classification
fitz-gov
pyrrho
text-embeddings-inference
Instructions to use yafitzdev/pyrrho-v1-nano-g2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use yafitzdev/pyrrho-v1-nano-g2 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="yafitzdev/pyrrho-v1-nano-g2")# Load model directly from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("yafitzdev/pyrrho-v1-nano-g2") model = AutoModelForSequenceClassification.from_pretrained("yafitzdev/pyrrho-v1-nano-g2") - Notebooks
- Google Colab
- Kaggle
Update v1 model card names
Browse files
README.md
CHANGED
|
@@ -19,9 +19,11 @@ metrics:
|
|
| 19 |
- f1
|
| 20 |
---
|
| 21 |
|
| 22 |
-
|
|
|
|
|
|
|
| 23 |
|
| 24 |
-
pyrrho-nano-g2 is a small RAG governance co-processor for anti-hallucination pipelines. It reads a user question plus retrieved source passages and returns an evidence-state decision a RAG application can use before answering: `ABSTAIN`, `DISPUTED`, or `TRUSTWORTHY`.
|
| 25 |
|
| 26 |
It is not an answer generator and not an open-world fact checker. It sits between retrieval and generation, or beside a generator as a fast guardrail, to reduce cases where unsupported or contradictory retrieved evidence gets treated as safe to answer from.
|
| 27 |
|
|
@@ -95,7 +97,7 @@ This model is not intended to write answers, verify facts outside the provided s
|
|
| 95 |
from transformers import AutoModelForSequenceClassification, AutoTokenizer
|
| 96 |
import torch
|
| 97 |
|
| 98 |
-
MODEL_ID = "yafitzdev/pyrrho-nano-g2"
|
| 99 |
LABELS = ["ABSTAIN", "DISPUTED", "TRUSTWORTHY"]
|
| 100 |
TAU = 0.50
|
| 101 |
|
|
@@ -150,7 +152,7 @@ from transformers import AutoTokenizer
|
|
| 150 |
import numpy as np
|
| 151 |
import onnxruntime as ort
|
| 152 |
|
| 153 |
-
MODEL_ID = "yafitzdev/pyrrho-nano-g2"
|
| 154 |
query = "Has the company achieved profitability?"
|
| 155 |
contexts = [
|
| 156 |
"The company posted its first profitable quarter, with net income of $4 million.",
|
|
@@ -263,10 +265,10 @@ The validation split was used for checkpoint and threshold selection. The held-o
|
|
| 263 |
|
| 264 |
```bibtex
|
| 265 |
@misc{pyrrho_nano_g2_2026,
|
| 266 |
-
title = {pyrrho-nano-g2},
|
| 267 |
author = {Yan Fitzner},
|
| 268 |
year = {2026},
|
| 269 |
-
url = {https://huggingface.co/yafitzdev/pyrrho-nano-g2},
|
| 270 |
}
|
| 271 |
```
|
| 272 |
|
|
|
|
| 19 |
- f1
|
| 20 |
---
|
| 21 |
|
| 22 |
+
> Historical Pyrrho v1 release. Current v2 model: [yafitzdev/pyrrho-v2-nano-g1](https://huggingface.co/yafitzdev/pyrrho-v2-nano-g1).
|
| 23 |
+
|
| 24 |
+
# pyrrho-v1-nano-g2
|
| 25 |
|
| 26 |
+
pyrrho-v1-nano-g2 is a small RAG governance co-processor for anti-hallucination pipelines. It reads a user question plus retrieved source passages and returns an evidence-state decision a RAG application can use before answering: `ABSTAIN`, `DISPUTED`, or `TRUSTWORTHY`.
|
| 27 |
|
| 28 |
It is not an answer generator and not an open-world fact checker. It sits between retrieval and generation, or beside a generator as a fast guardrail, to reduce cases where unsupported or contradictory retrieved evidence gets treated as safe to answer from.
|
| 29 |
|
|
|
|
| 97 |
from transformers import AutoModelForSequenceClassification, AutoTokenizer
|
| 98 |
import torch
|
| 99 |
|
| 100 |
+
MODEL_ID = "yafitzdev/pyrrho-v1-nano-g2"
|
| 101 |
LABELS = ["ABSTAIN", "DISPUTED", "TRUSTWORTHY"]
|
| 102 |
TAU = 0.50
|
| 103 |
|
|
|
|
| 152 |
import numpy as np
|
| 153 |
import onnxruntime as ort
|
| 154 |
|
| 155 |
+
MODEL_ID = "yafitzdev/pyrrho-v1-nano-g2"
|
| 156 |
query = "Has the company achieved profitability?"
|
| 157 |
contexts = [
|
| 158 |
"The company posted its first profitable quarter, with net income of $4 million.",
|
|
|
|
| 265 |
|
| 266 |
```bibtex
|
| 267 |
@misc{pyrrho_nano_g2_2026,
|
| 268 |
+
title = {pyrrho-v1-nano-g2},
|
| 269 |
author = {Yan Fitzner},
|
| 270 |
year = {2026},
|
| 271 |
+
url = {https://huggingface.co/yafitzdev/pyrrho-v1-nano-g2},
|
| 272 |
}
|
| 273 |
```
|
| 274 |
|