Update README.md
Browse files
README.md
CHANGED
|
@@ -44,7 +44,7 @@ As a result, Klear-46B-A2.5B-Base matches or surpasses the performance of dense
|
|
| 44 |
|
| 45 |
The base and instruction tuned + DPO models have the following architecture:
|
| 46 |
|
| 47 |
-
| **
|
| 48 |
|---------------------------|------------------------------------------------------------------------|
|
| 49 |
| hidden_size | 2048 |
|
| 50 |
| moe_intermediate_size | 896 |
|
|
|
|
| 44 |
|
| 45 |
The base and instruction tuned + DPO models have the following architecture:
|
| 46 |
|
| 47 |
+
| **key** | **value** |
|
| 48 |
|---------------------------|------------------------------------------------------------------------|
|
| 49 |
| hidden_size | 2048 |
|
| 50 |
| moe_intermediate_size | 896 |
|