Update README.md
Browse files
README.md
CHANGED
|
@@ -27,6 +27,7 @@ They're simple to use, fast and very accurate.
|
|
| 27 |
Like many people, I was excited about the architecture and training uplift from the ModernBERT architecture (`answerdotai/ModernBERT-base`).
|
| 28 |
So I've applied it to the stsb cross encoder, which is a very handy model. Additionally, I've added
|
| 29 |
pretraining from my much larger semi-synthetic dataset `dleemiller/wiki-sim` that targets this kind of objective.
|
|
|
|
| 30 |
|
| 31 |
---
|
| 32 |
|
|
|
|
| 27 |
Like many people, I was excited about the architecture and training uplift from the ModernBERT architecture (`answerdotai/ModernBERT-base`).
|
| 28 |
So I've applied it to the stsb cross encoder, which is a very handy model. Additionally, I've added
|
| 29 |
pretraining from my much larger semi-synthetic dataset `dleemiller/wiki-sim` that targets this kind of objective.
|
| 30 |
+
The inference performance efficiency, expanded context and simplicity make this a really nice platform as an evaluator model.
|
| 31 |
|
| 32 |
---
|
| 33 |
|