Files
Stack-3.0-Omni-Nexus/benchmarks/winogrande.json

8 lines
190 B
JSON
Raw Permalink Normal View History

{
"benchmark": "winogrande",
"model": "omni-nexus-alpha-q8",
"method": "chat-api (fill-blank, option word count)",
"accuracy": 0.5201262825572218,
"correct": 659,
"total": 1267
}