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

{
    "call": {
        "memory": {
            "unit": "MB",
            "max_ram": 1958.27712,
            "max_global_vram": 8311.025664,
            "max_process_vram": 1667236.630528,
            "max_reserved": 7864.32,
            "max_allocated": 7784.654848
        },
        "latency": {
            "unit": "s",
            "count": 3,
            "total": 1.4523962097167968,
            "mean": 0.4841320699055989,
            "stdev": 0.0027081496363248335,
            "p50": 0.48275347900390625,
            "p90": 0.48688336181640623,
            "p95": 0.48739959716796877,
            "p99": 0.48781258544921874,
            "values": [
                0.48791583251953125,
                0.48275347900390625,
                0.4817268981933594
            ]
        },
        "throughput": {
            "unit": "images/s",
            "value": 2.065552071762134
        },
        "energy": null,
        "efficiency": null
    }
}