Update README.md
Browse files
README.md
CHANGED
|
@@ -91,7 +91,6 @@ SABER utilizes a rigorous two-stage optimization pipeline: first, we adapted **M
|
|
| 91 |
<img src="https://cdn-uploads.huggingface.co/production/uploads/628f7a71dd993507cfcbe587/zC5JCPbsnIz-jflmTWae8.png" alt="SABER Training Pipeline" width="500"/>
|
| 92 |
</div>
|
| 93 |
|
| 94 |
-
|
| 95 |
---
|
| 96 |
|
| 97 |
**SABER is designed for:**
|
|
@@ -140,7 +139,7 @@ This model is the result of extensive evaluation across STS, triplets, and domai
|
|
| 140 |
|
| 141 |
## 📚 Training Data
|
| 142 |
|
| 143 |
-
SABER was trained on
|
| 144 |
|
| 145 |
* **2964 triplets** (Anchor, Positive, Negative)
|
| 146 |
* **21 domains**, including:
|
|
|
|
| 91 |
<img src="https://cdn-uploads.huggingface.co/production/uploads/628f7a71dd993507cfcbe587/zC5JCPbsnIz-jflmTWae8.png" alt="SABER Training Pipeline" width="500"/>
|
| 92 |
</div>
|
| 93 |
|
|
|
|
| 94 |
---
|
| 95 |
|
| 96 |
**SABER is designed for:**
|
|
|
|
| 139 |
|
| 140 |
## 📚 Training Data
|
| 141 |
|
| 142 |
+
**SABER** was trained on [Omartificial-Intelligence-Space/SaudiDialect-Triplet-21](https://huggingface.co/datasets/Omartificial-Intelligence-Space/SaudiDialect-Triplet-21), which contains:
|
| 143 |
|
| 144 |
* **2964 triplets** (Anchor, Positive, Negative)
|
| 145 |
* **21 domains**, including:
|