Add Terminal-Bench 2.0 evaluation results

This PR ensures your model shows up at https://huggingface.co/datasets/harborframework/terminal-bench-2.0.

This is based on a new evaluation results feature: https://huggingface.co/docs/hub/eval-results.

.eval_results/terminal_bench_2.yaml

@@ -0,0 +1,9 @@
+- dataset:
+    id: harborframework/terminal-bench-2.0
+    task_id: terminalbench_2
+  value: 20.2
+  source:
+    url: https://arxiv.org/pdf/2602.21193
+    name: Nemotron-Terminal-14B technical report
+    user: nielsr
+  notes: "agent: Terminus 2"