Evaluation Results

 This PR adds evaluation results extracted from the Model Card.

 **Benchmarks:**
 - MMLU-Pro: 87.8

• HLE: 28.7

• HLE: 48.3

   **Files created:**
   - .eval_results/mmlu_pro.yaml
  

• .eval_results/hle.yaml

• .eval_results/hle_with_tools.yaml

   ---
  
   Extracted automatically using the [LLM-powered evaluation extractor](https://github.com/huggingface/community-evals).
  

Ssss