120 lines
4.5 KiB
JSON
120 lines
4.5 KiB
JSON
{
|
|
"name": "qwen3-max@iquiz",
|
|
"dataset_name": "iquiz",
|
|
"dataset_pretty_name": "IQuiz",
|
|
"dataset_description": "\n## Overview\n\nIQuiz is a Chinese benchmark for evaluating AI models on intelligence quotient (IQ) and emotional quotient (EQ) questions. It tests logical reasoning, pattern recognition, and social-emotional understanding through multiple-choice questions.\n\n## Task Description\n\n- **Task Type**: Multiple-Choice Question Answering\n- **Input**: Question in Chinese with multiple choice options\n- **Output**: Selected answer with explanation (Chain-of-Thought)\n- **Language**: Chinese\n\n## Key Features\n\n- Dual evaluation of IQ and EQ capabilities\n- Chinese-language cognitive assessment\n- Multiple difficulty levels\n- Requires explanation alongside answer selection\n- Tests logical reasoning and emotional understanding\n\n## Evaluation Notes\n\n- Default configuration uses **0-shot** evaluation\n- Primary metric: **Accuracy**\n- Subsets: **IQ** (logical reasoning) and **EQ** (emotional intelligence)\n- Uses Chinese Chain-of-Thought prompt template\n- Evaluates on **test** split\n- Metadata includes difficulty level information\n",
|
|
"model_name": "qwen3-max",
|
|
"score": 0.8,
|
|
"metrics": [
|
|
{
|
|
"name": "mean_acc",
|
|
"num": 20,
|
|
"score": 0.8,
|
|
"macro_score": 0.8,
|
|
"categories": [
|
|
{
|
|
"name": [
|
|
"default"
|
|
],
|
|
"num": 20,
|
|
"score": 0.8,
|
|
"macro_score": 0.8,
|
|
"subsets": [
|
|
{
|
|
"name": "IQ",
|
|
"score": 1.0,
|
|
"num": 10
|
|
},
|
|
{
|
|
"name": "EQ",
|
|
"score": 0.6,
|
|
"num": 10
|
|
}
|
|
]
|
|
}
|
|
]
|
|
}
|
|
],
|
|
"analysis": "N/A",
|
|
"perf_metrics": {
|
|
"summary": {
|
|
"n_samples": 20,
|
|
"latency": {
|
|
"mean": 10.509281,
|
|
"std": 3.593791,
|
|
"min": 4.535925,
|
|
"25%": 8.009143,
|
|
"50%": 10.797033,
|
|
"75%": 12.824467,
|
|
"90%": 14.000725,
|
|
"99%": 17.871054,
|
|
"max": 18.672862
|
|
},
|
|
"throughput": {
|
|
"avg_output_tps": 32.05,
|
|
"avg_req_ps": 0.0952
|
|
},
|
|
"usage": {
|
|
"input_tokens": {
|
|
"mean": 152.35,
|
|
"std": 53.29686,
|
|
"min": 98.0,
|
|
"25%": 109.0,
|
|
"50%": 138.0,
|
|
"75%": 180.5,
|
|
"90%": 232.0,
|
|
"99%": 264.58,
|
|
"max": 268.0
|
|
},
|
|
"output_tokens": {
|
|
"mean": 336.85,
|
|
"std": 121.738838,
|
|
"min": 137.0,
|
|
"25%": 250.25,
|
|
"50%": 341.5,
|
|
"75%": 415.25,
|
|
"90%": 462.7,
|
|
"99%": 579.97,
|
|
"max": 606.0
|
|
},
|
|
"total_tokens": {
|
|
"mean": 489.2,
|
|
"std": 166.081654,
|
|
"min": 249.0,
|
|
"25%": 358.25,
|
|
"50%": 498.0,
|
|
"75%": 591.5,
|
|
"90%": 684.1,
|
|
"99%": 843.22,
|
|
"max": 874.0
|
|
},
|
|
"total_input_tokens": 3047,
|
|
"total_output_tokens": 6737,
|
|
"total_tokens_count": 9784
|
|
},
|
|
"ttft": {
|
|
"mean": 0.576507,
|
|
"std": 0.133631,
|
|
"min": 0.395713,
|
|
"25%": 0.503878,
|
|
"50%": 0.575535,
|
|
"75%": 0.649104,
|
|
"90%": 0.677137,
|
|
"99%": 0.926158,
|
|
"max": 0.983595
|
|
},
|
|
"tpot": {
|
|
"mean": 0.029563,
|
|
"std": 0.000272,
|
|
"min": 0.028988,
|
|
"25%": 0.029437,
|
|
"50%": 0.02956,
|
|
"75%": 0.029705,
|
|
"90%": 0.029903,
|
|
"99%": 0.030089,
|
|
"max": 0.030124
|
|
}
|
|
}
|
|
},
|
|
"num": 20
|
|
} |