SPX-CD-Pro / evaluation_results.json
Surd-AI's picture
Initial model release
04f53c5
Raw History Blame Contribute Delete
1.51 kB
{
"benchmark": "CMDB-1500",
"adapter_sha256": "2982cdb8e60cec22b45da33720cd7b397381d910a12d4773be199493e2b400a4",
"backend": "vLLM",
"base_precision": "AWQ INT4",
"effort": 2,
"full": {
"n": 1500,
"valid": 1500,
"accuracy_percent": 81.2
},
"text": {
"n": 1200,
"accuracy_percent": 79.33333333333333
},
"images": {
"n": 300,
"accuracy_percent": 88.66666666666667
},
"provenance": "CMDB-1500 results for the adapter SHA and configuration specified here.",
"website_reported_evaluations": {
"source_url": "https://surdai.com/zh/blog/simplex-cd-release#cd-latest-results",
"checked_date": "2026-10-09",
"provenance": "Results reported by the linked SurdAI release page using its evaluation serving configurations.",
"jevbench_public231": [
{
"effort": 1,
"n": 231,
"correct": 207,
"accuracy_percent": 89.6103896103896
},
{
"effort": 2,
"n": 231,
"correct": 208,
"accuracy_percent": 90.04329004329004
}
],
"jevbench_hard111": [
{
"effort": 1,
"n": 111,
"correct": 88,
"accuracy_percent": 79.27927927927928
},
{
"effort": 2,
"n": 111,
"correct": 89,
"accuracy_percent": 80.18018018018019
}
],
"hard_is_subset_of_public": true,
"decision_index": {
"edition": "0.2.1",
"effort": 1,
"score": 58.3
},
"ece_imported": false
}
}