Upload benchmark_results.json with huggingface_hub

benchmark_results.json

@@ -0,0 +1,59 @@
+{
+  "timestamp": "2026-06-09T08:49:09.527775",
+  "models": [
+    {
+      "model": "DeepSeek V4 Flash",
+      "id": "deepseek/deepseek-v4-flash",
+      "mmlu": {
+        "score": 100.0,
+        "correct": 3,
+        "total": 3,
+        "avg_latency": 0.99
+      },
+      "math": {
+        "score": 100.0,
+        "correct": 2,
+        "total": 2,
+        "avg_latency": 3.25
+      },
+      "bfcl": {
+        "score": 100.0,
+        "correct": 2,
+        "total": 2
+      },
+      "humaneval": {
+        "score": 100.0,
+        "correct": 2,
+        "total": 2
+      },
+      "overall": 100.0
+    },
+    {
+      "model": "Claude Sonnet 4",
+      "id": "anthropic/claude-sonnet-4",
+      "mmlu": {
+        "score": 100.0,
+        "correct": 3,
+        "total": 3,
+        "avg_latency": 1.12
+      },
+      "math": {
+        "score": 50.0,
+        "correct": 1,
+        "total": 2,
+        "avg_latency": 4.18
+      },
+      "bfcl": {
+        "score": 100.0,
+        "correct": 2,
+        "total": 2
+      },
+      "humaneval": {
+        "score": 100.0,
+        "correct": 2,
+        "total": 2
+      },
+      "overall": 87.5
+    }
+  ]
+}