cuda_inference_diffusers_text-to-image_CompVis/stable-diffusion-v1-4/benchmark_report.json

{
    "load_model": {
        "memory": {
            "unit": "MB",
            "max_ram": 4560.379904,
            "max_global_vram": 68702.69952,
            "max_process_vram": 453057.466368,
            "max_reserved": 5567.93856,
            "max_allocated": 5510.160384
        },
        "latency": {
            "unit": "s",
            "values": [
                4.3272001953125
            ],
            "count": 1,
            "total": 4.3272001953125,
            "mean": 4.3272001953125,
            "p50": 4.3272001953125,
            "p90": 4.3272001953125,
            "p95": 4.3272001953125,
            "p99": 4.3272001953125,
            "stdev": 0,
            "stdev_": 0
        },
        "throughput": null,
        "energy": null,
        "efficiency": null
    },
    "call": {
        "memory": {
            "unit": "MB",
            "max_ram": 2110.373888,
            "max_global_vram": 68702.69952,
            "max_process_vram": 3500357.779456,
            "max_reserved": 10137.632768,
            "max_allocated": 10060.197376
        },
        "latency": {
            "unit": "s",
            "values": [
                0.8474823608398437,
                0.84228369140625
            ],
            "count": 2,
            "total": 1.6897660522460938,
            "mean": 0.8448830261230469,
            "p50": 0.8448830261230469,
            "p90": 0.8469624938964844,
            "p95": 0.847222427368164,
            "p99": 0.8474303741455078,
            "stdev": 0.002599334716796853,
            "stdev_": 0.3076561649870679
        },
        "throughput": {
            "unit": "images/s",
            "value": 2.367191597134447
        },
        "energy": null,
        "efficiency": null
    }
}