Update README.md

README.md

@@ -90,17 +90,17 @@ We recommend fine-tuning LFM2.5 for your specific use case to achieve the best r
 
 ## 📊 Performance
 
-| Model | [JMMLU](https://arxiv.org/pdf/2402.14531) | [M-IFEval (ja)](https://arxiv.org/pdf/2502.04688) |
-|-------|------|----------|
-| **LFM2.5-1.2B-JP** | 49.8 | 60.6 | 
-| **[LFM2.5-1.2B-Instruct](https://huggingface.co/LiquidAI/LFM2.5-1.2B-Instruct)** | 47.5 | 41.8 |
-| Qwen3-1.7B (Instruct mode) | 47.1 | 40.3 | 
-| Llama 3.2 1B Instruct | 33.8 | 24.1 |
-| TinySwallow-1.5B-Instruct | 48.0 | 36.5 |
-| Gemma-2-Llama-Swallow-2b-it-v0.1 | 48.1 | 33.4 |
-| Gemma-3-1b-it | 34.5 | 26.3 |
-| Granite-4.0-h-1b | 42.2 | 39.3 |
-| Sarashina2.2-1b-instruct-v0.1 | 40.2 | 21.9 |
+| Model | [JMMLU](https://arxiv.org/pdf/2402.14531) | [M-IFEval (ja)](https://arxiv.org/pdf/2502.04688) | [GSM8K (ja)](https://huggingface.co/datasets/SakanaAI/gsm8k-ja-test_250-1319) | 
+|-------|------|----------|----------|
+| **LFM2.5-1.2B-JP** | 49.8 | 60.6 | TBD |
+| **[LFM2.5-1.2B-Instruct](https://huggingface.co/LiquidAI/LFM2.5-1.2B-Instruct)** | 47.5 | 41.8 | 46.8 |
+| Qwen3-1.7B (Instruct mode) | 47.1 | 40.3 | 46.0 |
+| Llama 3.2 1B Instruct | 33.8 | 24.1 | 25.2 |
+| TinySwallow-1.5B-Instruct | 48.0 | 36.5 | 47.2 |
+| Gemma-2-Llama-Swallow-2b-it-v0.1 | 48.1 | 33.4 | 34.4 |
+| Gemma-3-1b-it | 34.5 | 26.3 | 33.6 |
+| Granite-4.0-h-1b | 42.2 | 39.3 | 42.8 |
+| Sarashina2.2-1b-instruct-v0.1 | 40.2 | 21.9 | 44.4 |
 
 **Evaluation Notes**