Update README.md
Browse files
README.md
CHANGED
|
@@ -51,7 +51,6 @@ exclude_layers = [
|
|
| 51 |
"*mlp.gate_proj",
|
| 52 |
"*mlp.up_proj",
|
| 53 |
"*mlp.down_proj",
|
| 54 |
-
"*shared_experts*",
|
| 55 |
]
|
| 56 |
|
| 57 |
# --- Build quant config from template ---
|
|
@@ -93,11 +92,11 @@ The model was evaluated on GSM8K benchmarks.
|
|
| 93 |
<tr>
|
| 94 |
<td>GSM8K (flexible-extract)
|
| 95 |
</td>
|
| 96 |
-
<td>95.45
|
| 97 |
-
</td>
|
| 98 |
<td>95.00
|
| 99 |
</td>
|
| 100 |
-
<td>
|
|
|
|
|
|
|
| 101 |
</td>
|
| 102 |
</tr>
|
| 103 |
</table>
|
|
|
|
| 51 |
"*mlp.gate_proj",
|
| 52 |
"*mlp.up_proj",
|
| 53 |
"*mlp.down_proj",
|
|
|
|
| 54 |
]
|
| 55 |
|
| 56 |
# --- Build quant config from template ---
|
|
|
|
| 92 |
<tr>
|
| 93 |
<td>GSM8K (flexible-extract)
|
| 94 |
</td>
|
|
|
|
|
|
|
| 95 |
<td>95.00
|
| 96 |
</td>
|
| 97 |
+
<td>94.92
|
| 98 |
+
</td>
|
| 99 |
+
<td>99.92%
|
| 100 |
</td>
|
| 101 |
</tr>
|
| 102 |
</table>
|