Add GPT-OSS-120B results for all 3 benchmarks

results/gpt-oss-120b-rh-swe-bench-claude-code.json

@@ -0,0 +1,48 @@
+{
+    "benchmark": {
+        "name": "RH SWE-Bench",
+        "repo": "rounakbende10/rh-swe-bench",
+        "num_tasks": 357,
+        "url": "https://huggingface.co/datasets/rounakbende/rh-swe-bench"
+    },
+    "harness": {
+        "name": "Claude Code",
+        "skills": [],
+        "is_oss": false,
+        "url": "https://docs.anthropic.com/en/docs/claude-code"
+    },
+    "model": {
+        "name": "GPT-OSS-120B",
+        "repo": "RedHatAI/gpt-oss-120b",
+        "is_oss": true,
+        "num_params": 120,
+        "precision": "mxfp4",
+        "url": "https://huggingface.co/RedHatAI/gpt-oss-120b"
+    },
+    "environment": {
+        "name": "harbor",
+        "config": {
+            "name": "rh-swe-bench"
+        },
+        "url": "https://github.com/harbor-framework/harbor"
+    },
+    "metrics": {
+        "n_tasks": 357,
+        "n_errors": 31,
+        "score": 0.317,
+        "n_input_tokens": 1088634865,
+        "n_cache_tokens": 0,
+        "n_output_tokens": 6200414,
+        "n_total_tokens": 1094835279,
+        "agent_time_seconds": 398444,
+        "total_time_seconds": 575536,
+        "cost_usd": 15.05,
+        "mean_input_tokens_per_task": 3049397,
+        "mean_cache_tokens_per_task": 0,
+        "mean_output_tokens_per_task": 17368,
+        "mean_tokens_per_task": 3066765,
+        "mean_cost_usd_per_task": 0.04,
+        "mean_total_time_seconds_per_task": 1612,
+        "mean_agent_time_seconds_per_task": 1116
+    }
+}

results/gpt-oss-120b-rh-swe-bench-opencode.json

@@ -0,0 +1,48 @@
+{
+    "benchmark": {
+        "name": "RH SWE-Bench",
+        "repo": "rounakbende10/rh-swe-bench",
+        "num_tasks": 357,
+        "url": "https://huggingface.co/datasets/rounakbende/rh-swe-bench"
+    },
+    "harness": {
+        "name": "OpenCode",
+        "skills": [],
+        "is_oss": true,
+        "url": "https://github.com/opencode-ai/opencode"
+    },
+    "model": {
+        "name": "GPT-OSS-120B",
+        "repo": "RedHatAI/gpt-oss-120b",
+        "is_oss": true,
+        "num_params": 120,
+        "precision": "mxfp4",
+        "url": "https://huggingface.co/RedHatAI/gpt-oss-120b"
+    },
+    "environment": {
+        "name": "harbor",
+        "config": {
+            "name": "rh-swe-bench"
+        },
+        "url": "https://github.com/harbor-framework/harbor"
+    },
+    "metrics": {
+        "n_tasks": 357,
+        "n_errors": 7,
+        "score": 0.294,
+        "n_input_tokens": 419671770,
+        "n_cache_tokens": 0,
+        "n_output_tokens": 2277963,
+        "n_total_tokens": 421949733,
+        "agent_time_seconds": 98909,
+        "total_time_seconds": 169021,
+        "cost_usd": 3.74,
+        "mean_input_tokens_per_task": 1175551,
+        "mean_cache_tokens_per_task": 0,
+        "mean_output_tokens_per_task": 6380,
+        "mean_tokens_per_task": 1181932,
+        "mean_cost_usd_per_task": 0.01,
+        "mean_total_time_seconds_per_task": 473,
+        "mean_agent_time_seconds_per_task": 277
+    }
+}

results/gpt-oss-120b-rh-swe-bench-pi.json

@@ -0,0 +1,48 @@
+{
+    "benchmark": {
+        "name": "RH SWE-Bench",
+        "repo": "rounakbende10/rh-swe-bench",
+        "num_tasks": 357,
+        "url": "https://huggingface.co/datasets/rounakbende/rh-swe-bench"
+    },
+    "harness": {
+        "name": "Pi",
+        "skills": [],
+        "is_oss": true,
+        "url": "https://github.com/earendil-works/pi/tree/main"
+    },
+    "model": {
+        "name": "GPT-OSS-120B",
+        "repo": "RedHatAI/gpt-oss-120b",
+        "is_oss": true,
+        "num_params": 120,
+        "precision": "mxfp4",
+        "url": "https://huggingface.co/RedHatAI/gpt-oss-120b"
+    },
+    "environment": {
+        "name": "harbor",
+        "config": {
+            "name": "rh-swe-bench"
+        },
+        "url": "https://github.com/harbor-framework/harbor"
+    },
+    "metrics": {
+        "n_tasks": 357,
+        "n_errors": 20,
+        "score": 0.23,
+        "n_input_tokens": 419282595,
+        "n_cache_tokens": 0,
+        "n_output_tokens": 1954817,
+        "n_total_tokens": 421237412,
+        "agent_time_seconds": 100632,
+        "total_time_seconds": 168697,
+        "cost_usd": 3.8,
+        "mean_input_tokens_per_task": 1174461,
+        "mean_cache_tokens_per_task": 0,
+        "mean_output_tokens_per_task": 5475,
+        "mean_tokens_per_task": 1179936,
+        "mean_cost_usd_per_task": 0.01,
+        "mean_total_time_seconds_per_task": 472,
+        "mean_agent_time_seconds_per_task": 281
+    }
+}

results/gpt-oss-120b-swebench-pro-ansible-claude-code.json

@@ -0,0 +1,48 @@
+{
+    "benchmark": {
+        "name": "SWE-Bench Pro",
+        "repo": "scale-ai/swe-bench-pro",
+        "num_tasks": 96,
+        "url": "https://huggingface.co/datasets/scale-ai/swe-bench-pro"
+    },
+    "harness": {
+        "name": "Claude Code",
+        "skills": [],
+        "is_oss": false,
+        "url": "https://docs.anthropic.com/en/docs/claude-code"
+    },
+    "model": {
+        "name": "GPT-OSS-120B",
+        "repo": "RedHatAI/gpt-oss-120b",
+        "is_oss": true,
+        "num_params": 120,
+        "precision": "mxfp4",
+        "url": "https://huggingface.co/RedHatAI/gpt-oss-120b"
+    },
+    "environment": {
+        "name": "harbor",
+        "config": {
+            "name": "swebench-pro-ansible"
+        },
+        "url": "https://github.com/harbor-framework/harbor"
+    },
+    "metrics": {
+        "n_tasks": 96,
+        "n_errors": 2,
+        "score": 0.375,
+        "n_input_tokens": 347744043,
+        "n_cache_tokens": 0,
+        "n_output_tokens": 1903023,
+        "n_total_tokens": 349647066,
+        "agent_time_seconds": 59653,
+        "total_time_seconds": 67586,
+        "cost_usd": 2.25,
+        "mean_input_tokens_per_task": 3622333,
+        "mean_cache_tokens_per_task": 0,
+        "mean_output_tokens_per_task": 19823,
+        "mean_tokens_per_task": 3642156,
+        "mean_cost_usd_per_task": 0.02,
+        "mean_total_time_seconds_per_task": 704,
+        "mean_agent_time_seconds_per_task": 621
+    }
+}

results/gpt-oss-120b-swebench-pro-ansible-opencode.json

@@ -0,0 +1,48 @@
+{
+    "benchmark": {
+        "name": "SWE-Bench Pro",
+        "repo": "scale-ai/swe-bench-pro",
+        "num_tasks": 96,
+        "url": "https://huggingface.co/datasets/scale-ai/swe-bench-pro"
+    },
+    "harness": {
+        "name": "OpenCode",
+        "skills": [],
+        "is_oss": true,
+        "url": "https://github.com/opencode-ai/opencode"
+    },
+    "model": {
+        "name": "GPT-OSS-120B",
+        "repo": "RedHatAI/gpt-oss-120b",
+        "is_oss": true,
+        "num_params": 120,
+        "precision": "mxfp4",
+        "url": "https://huggingface.co/RedHatAI/gpt-oss-120b"
+    },
+    "environment": {
+        "name": "harbor",
+        "config": {
+            "name": "swebench-pro-ansible"
+        },
+        "url": "https://github.com/harbor-framework/harbor"
+    },
+    "metrics": {
+        "n_tasks": 96,
+        "n_errors": 1,
+        "score": 0.333,
+        "n_input_tokens": 112041538,
+        "n_cache_tokens": 0,
+        "n_output_tokens": 781173,
+        "n_total_tokens": 112822711,
+        "agent_time_seconds": 25086,
+        "total_time_seconds": 31180,
+        "cost_usd": 0.95,
+        "mean_input_tokens_per_task": 1167099,
+        "mean_cache_tokens_per_task": 0,
+        "mean_output_tokens_per_task": 8137,
+        "mean_tokens_per_task": 1175236,
+        "mean_cost_usd_per_task": 0.01,
+        "mean_total_time_seconds_per_task": 324,
+        "mean_agent_time_seconds_per_task": 261
+    }
+}

results/gpt-oss-120b-swebench-pro-ansible-pi.json

@@ -0,0 +1,48 @@
+{
+    "benchmark": {
+        "name": "SWE-Bench Pro",
+        "repo": "scale-ai/swe-bench-pro",
+        "num_tasks": 96,
+        "url": "https://huggingface.co/datasets/scale-ai/swe-bench-pro"
+    },
+    "harness": {
+        "name": "Pi",
+        "skills": [],
+        "is_oss": true,
+        "url": "https://github.com/earendil-works/pi/tree/main"
+    },
+    "model": {
+        "name": "GPT-OSS-120B",
+        "repo": "RedHatAI/gpt-oss-120b",
+        "is_oss": true,
+        "num_params": 120,
+        "precision": "mxfp4",
+        "url": "https://huggingface.co/RedHatAI/gpt-oss-120b"
+    },
+    "environment": {
+        "name": "harbor",
+        "config": {
+            "name": "swebench-pro-ansible"
+        },
+        "url": "https://github.com/harbor-framework/harbor"
+    },
+    "metrics": {
+        "n_tasks": 96,
+        "n_errors": 15,
+        "score": 0.292,
+        "n_input_tokens": 125247664,
+        "n_cache_tokens": 0,
+        "n_output_tokens": 597435,
+        "n_total_tokens": 125845099,
+        "agent_time_seconds": 75715,
+        "total_time_seconds": 79741,
+        "cost_usd": 2.86,
+        "mean_input_tokens_per_task": 1304663,
+        "mean_cache_tokens_per_task": 0,
+        "mean_output_tokens_per_task": 6223,
+        "mean_tokens_per_task": 1310886,
+        "mean_cost_usd_per_task": 0.03,
+        "mean_total_time_seconds_per_task": 830,
+        "mean_agent_time_seconds_per_task": 788
+    }
+}

results/gpt-oss-120b-swebench-verified-claude-code.json

@@ -0,0 +1,48 @@
+{
+    "benchmark": {
+        "name": "SWE-Bench Verified",
+        "repo": "swe-bench/swe-bench-verified",
+        "num_tasks": 500,
+        "url": "https://huggingface.co/datasets/princeton-nlp/SWE-bench_Verified"
+    },
+    "harness": {
+        "name": "Claude Code",
+        "skills": [],
+        "is_oss": false,
+        "url": "https://docs.anthropic.com/en/docs/claude-code"
+    },
+    "model": {
+        "name": "GPT-OSS-120B",
+        "repo": "RedHatAI/gpt-oss-120b",
+        "is_oss": true,
+        "num_params": 120,
+        "precision": "mxfp4",
+        "url": "https://huggingface.co/RedHatAI/gpt-oss-120b"
+    },
+    "environment": {
+        "name": "harbor",
+        "config": {
+            "name": "swebench-verified"
+        },
+        "url": "https://github.com/harbor-framework/harbor"
+    },
+    "metrics": {
+        "n_tasks": 500,
+        "n_errors": 13,
+        "score": 0.376,
+        "n_input_tokens": 833651681,
+        "n_cache_tokens": 0,
+        "n_output_tokens": 5656173,
+        "n_total_tokens": 839307854,
+        "agent_time_seconds": 150981,
+        "total_time_seconds": 200547,
+        "cost_usd": 5.7,
+        "mean_input_tokens_per_task": 1667303,
+        "mean_cache_tokens_per_task": 0,
+        "mean_output_tokens_per_task": 11312,
+        "mean_tokens_per_task": 1678615,
+        "mean_cost_usd_per_task": 0.01,
+        "mean_total_time_seconds_per_task": 401,
+        "mean_agent_time_seconds_per_task": 301
+    }
+}

results/gpt-oss-120b-swebench-verified-opencode.json

@@ -0,0 +1,48 @@
+{
+    "benchmark": {
+        "name": "SWE-Bench Verified",
+        "repo": "swe-bench/swe-bench-verified",
+        "num_tasks": 500,
+        "url": "https://huggingface.co/datasets/princeton-nlp/SWE-bench_Verified"
+    },
+    "harness": {
+        "name": "OpenCode",
+        "skills": [],
+        "is_oss": true,
+        "url": "https://github.com/opencode-ai/opencode"
+    },
+    "model": {
+        "name": "GPT-OSS-120B",
+        "repo": "RedHatAI/gpt-oss-120b",
+        "is_oss": true,
+        "num_params": 120,
+        "precision": "mxfp4",
+        "url": "https://huggingface.co/RedHatAI/gpt-oss-120b"
+    },
+    "environment": {
+        "name": "harbor",
+        "config": {
+            "name": "swebench-verified"
+        },
+        "url": "https://github.com/harbor-framework/harbor"
+    },
+    "metrics": {
+        "n_tasks": 500,
+        "n_errors": 33,
+        "score": 0.326,
+        "n_input_tokens": 202856940,
+        "n_cache_tokens": 0,
+        "n_output_tokens": 1444371,
+        "n_total_tokens": 204301311,
+        "agent_time_seconds": 35190,
+        "total_time_seconds": 322482,
+        "cost_usd": 3.32,
+        "mean_input_tokens_per_task": 405713,
+        "mean_cache_tokens_per_task": 0,
+        "mean_output_tokens_per_task": 2888,
+        "mean_tokens_per_task": 408602,
+        "mean_cost_usd_per_task": 0.01,
+        "mean_total_time_seconds_per_task": 644,
+        "mean_agent_time_seconds_per_task": 70
+    }
+}

results/gpt-oss-120b-swebench-verified-pi.json

@@ -0,0 +1,48 @@
+{
+    "benchmark": {
+        "name": "SWE-Bench Verified",
+        "repo": "swe-bench/swe-bench-verified",
+        "num_tasks": 500,
+        "url": "https://huggingface.co/datasets/princeton-nlp/SWE-bench_Verified"
+    },
+    "harness": {
+        "name": "Pi",
+        "skills": [],
+        "is_oss": true,
+        "url": "https://github.com/earendil-works/pi/tree/main"
+    },
+    "model": {
+        "name": "GPT-OSS-120B",
+        "repo": "RedHatAI/gpt-oss-120b",
+        "is_oss": true,
+        "num_params": 120,
+        "precision": "mxfp4",
+        "url": "https://huggingface.co/RedHatAI/gpt-oss-120b"
+    },
+    "environment": {
+        "name": "harbor",
+        "config": {
+            "name": "swebench-verified"
+        },
+        "url": "https://github.com/harbor-framework/harbor"
+    },
+    "metrics": {
+        "n_tasks": 500,
+        "n_errors": 47,
+        "score": 0.234,
+        "n_input_tokens": 90107247,
+        "n_cache_tokens": 0,
+        "n_output_tokens": 905031,
+        "n_total_tokens": 91012278,
+        "agent_time_seconds": 146747,
+        "total_time_seconds": 176174,
+        "cost_usd": 13.86,
+        "mean_input_tokens_per_task": 180214,
+        "mean_cache_tokens_per_task": 0,
+        "mean_output_tokens_per_task": 1810,
+        "mean_tokens_per_task": 182024,
+        "mean_cost_usd_per_task": 0.03,
+        "mean_total_time_seconds_per_task": 352,
+        "mean_agent_time_seconds_per_task": 293
+    }
+}