Update README.md
Browse files
README.md
CHANGED
|
@@ -45,10 +45,10 @@ FlagEval (Libra)** is a comprehensive evaluation system and open platform for la
|
|
| 45 |
|
| 46 |
| Metrics | Kimi-K2-Instruct-FlagOS-H100-CUDA | Kimi-K2-Instruct-FlagOS-FlagOS-Nvidia |
|
| 47 |
| --------- | -------------------------------- | ------------------------------------ |
|
| 48 |
-
| AIME | 0.667 | 0.700 |
|
| 49 |
-
| LiveBench | 0.685 | 0.690 |
|
| 50 |
-
| MMLU | 0.773 | 0.788 |
|
| 51 |
-
| MUSR | 0.724 | 0.710 |
|
| 52 |
|
| 53 |
# User Guide
|
| 54 |
|
|
|
|
| 45 |
|
| 46 |
| Metrics | Kimi-K2-Instruct-FlagOS-H100-CUDA | Kimi-K2-Instruct-FlagOS-FlagOS-Nvidia |
|
| 47 |
| --------- | -------------------------------- | ------------------------------------ |
|
| 48 |
+
| AIME-0shot@avg1 | 0.667 | 0.700 |
|
| 49 |
+
| LiveBench-0shot@avg1 | 0.685 | 0.690 |
|
| 50 |
+
| MMLU-0shot@avg1 | 0.773 | 0.788 |
|
| 51 |
+
| MUSR-5shots@avg1 | 0.724 | 0.710 |
|
| 52 |
|
| 53 |
# User Guide
|
| 54 |
|