Update README.md
Browse files
README.md
CHANGED
|
@@ -30,7 +30,7 @@ These models were chosen because `fblgit/UNA-TheBeagle-7b-v1` has excellent perf
|
|
| 30 |
|
| 31 |
## Benchmark
|
| 32 |
|
| 33 |
-
On a synthetic benchmark of
|
| 34 |
|
| 35 |
Nous Benchmark numbers shall be available soon.
|
| 36 |
|
|
|
|
| 30 |
|
| 31 |
## Benchmark
|
| 32 |
|
| 33 |
+
On a synthetic benchmark of 35 medical diagnosis questions generated by GPT-4, GPT-4 also being an evaluator, MedleyMD scored **96.25/100**.
|
| 34 |
|
| 35 |
Nous Benchmark numbers shall be available soon.
|
| 36 |
|