Update README.md
Browse files
README.md
CHANGED
|
@@ -4,7 +4,7 @@ tags:
|
|
| 4 |
- security
|
| 5 |
- blue-team
|
| 6 |
- roberta
|
| 7 |
-
license:
|
| 8 |
datasets:
|
| 9 |
- trendmicro-ailab/Primus-FineWeb
|
| 10 |
metrics:
|
|
@@ -44,8 +44,6 @@ Detects **blue-team / defensive security** text (English), with a focus on **tec
|
|
| 44 |
| Other | 130 000 |
|
| 45 |
| **Total** | **180 296** |
|
| 46 |
|
| 47 |
-
Source: [Primus-FineWeb](https://huggingface.co/datasets/trendmicro-ailab/Primus-FineWeb)
|
| 48 |
-
|
| 49 |
---
|
| 50 |
|
| 51 |
## Model details
|
|
@@ -60,6 +58,16 @@ Source: [Primus-FineWeb](https://huggingface.co/datasets/trendmicro-ailab/Primus
|
|
| 60 |
|
| 61 |
---
|
| 62 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 63 |
## Quick start
|
| 64 |
|
| 65 |
```python
|
|
@@ -75,4 +83,4 @@ prob = clf(text)[0]["score"] # sigmoid prob for class 0 (Defensive)
|
|
| 75 |
print(f"P(defensive) = {prob:.3f}")
|
| 76 |
|
| 77 |
is_blue = prob >= 0.579 # ← recommended threshold
|
| 78 |
-
print("is_blue:", is_blue)
|
|
|
|
| 4 |
- security
|
| 5 |
- blue-team
|
| 6 |
- roberta
|
| 7 |
+
license: odc-by
|
| 8 |
datasets:
|
| 9 |
- trendmicro-ailab/Primus-FineWeb
|
| 10 |
metrics:
|
|
|
|
| 44 |
| Other | 130 000 |
|
| 45 |
| **Total** | **180 296** |
|
| 46 |
|
|
|
|
|
|
|
| 47 |
---
|
| 48 |
|
| 49 |
## Model details
|
|
|
|
| 58 |
|
| 59 |
---
|
| 60 |
|
| 61 |
+
## Training Data License
|
| 62 |
+
|
| 63 |
+
- **Source**: [trendmicro-ailab/Primus-FineWeb](https://huggingface.co/datasets/trendmicro-ailab/Primus-FineWeb)
|
| 64 |
+
- **License**: ODC-By-1.0 (http://opendatacommons.org/licenses/by/1-0/)
|
| 65 |
+
- **Requirements**:
|
| 66 |
+
- Preserve all original copyright/license notices
|
| 67 |
+
- Honor [Common Crawl ToU](https://commoncrawl.org/terms-of-use/)
|
| 68 |
+
|
| 69 |
+
---
|
| 70 |
+
|
| 71 |
## Quick start
|
| 72 |
|
| 73 |
```python
|
|
|
|
| 83 |
print(f"P(defensive) = {prob:.3f}")
|
| 84 |
|
| 85 |
is_blue = prob >= 0.579 # ← recommended threshold
|
| 86 |
+
print("is_blue:", is_blue)
|