Update README.md
Browse files
README.md
CHANGED
|
@@ -8654,9 +8654,12 @@ language:
|
|
| 8654 |
---
|
| 8655 |
|
| 8656 |
# Mini-GTE
|
| 8657 |
-
## Overview
|
| 8658 |
|
| 8659 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8660 |
This is the first model developed by QTACK and serves as a proof of concept for our distillation approach! Built upon a distillbert-based architecture, Mini-GTE is distilled from GTE and designed for efficiency without sacrificing accuracy at only 66M parameters. As a standalone sentence transformer, it ranks 2nd on the MTEB classic leaderboard in the <100M parameter category and 63rd overall which makes it a strong choice for real-time query encoding, semantic search, and similarity tasks.
|
| 8661 |
|
| 8662 |
## Model Details
|
|
|
|
| 8654 |
---
|
| 8655 |
|
| 8656 |
# Mini-GTE
|
|
|
|
| 8657 |
|
| 8658 |
+
<p align="center">
|
| 8659 |
+
<img src="./qtack_logo.png" alt="QTACK Logo" style="width:33%;">
|
| 8660 |
+
</p>
|
| 8661 |
+
|
| 8662 |
+
## Overview
|
| 8663 |
This is the first model developed by QTACK and serves as a proof of concept for our distillation approach! Built upon a distillbert-based architecture, Mini-GTE is distilled from GTE and designed for efficiency without sacrificing accuracy at only 66M parameters. As a standalone sentence transformer, it ranks 2nd on the MTEB classic leaderboard in the <100M parameter category and 63rd overall which makes it a strong choice for real-time query encoding, semantic search, and similarity tasks.
|
| 8664 |
|
| 8665 |
## Model Details
|