cpu_inference_transformers_text-generation_openai-community/gpt2/benchmark.json

{
    "config": {
        "name": "cpu_inference_transformers_text-generation_openai-community/gpt2",
        "backend": {
            "name": "pytorch",
            "version": "2.3.0+cpu",
            "_target_": "optimum_benchmark.backends.pytorch.backend.PyTorchBackend",
            "task": "text-generation",
            "library": "transformers",
            "model": "openai-community/gpt2",
            "processor": "openai-community/gpt2",
            "device": "cpu",
            "device_ids": null,
            "seed": 42,
            "inter_op_num_threads": null,
            "intra_op_num_threads": null,
            "model_kwargs": {},
            "processor_kwargs": {},
            "hub_kwargs": {},
            "no_weights": true,
            "device_map": null,
            "torch_dtype": null,
            "eval_mode": true,
            "to_bettertransformer": false,
            "low_cpu_mem_usage": null,
            "attn_implementation": null,
            "cache_implementation": null,
            "autocast_enabled": false,
            "autocast_dtype": null,
            "torch_compile": false,
            "torch_compile_target": "forward",
            "torch_compile_config": {},
            "quantization_scheme": null,
            "quantization_config": {},
            "deepspeed_inference": false,
            "deepspeed_inference_config": {},
            "peft_type": null,
            "peft_config": {}
        },
        "scenario": {
            "name": "inference",
            "_target_": "optimum_benchmark.scenarios.inference.scenario.InferenceScenario",
            "iterations": 1,
            "duration": 1,
            "warmup_runs": 1,
            "input_shapes": {
                "batch_size": 1,
                "num_choices": 2,
                "sequence_length": 2
            },
            "new_tokens": null,
            "latency": true,
            "memory": true,
            "energy": true,
            "forward_kwargs": {},
            "generate_kwargs": {
                "max_new_tokens": 2,
                "min_new_tokens": 2
            },
            "call_kwargs": {
                "num_inference_steps": 2
            }
        },
        "launcher": {
            "name": "process",
            "_target_": "optimum_benchmark.launchers.process.launcher.ProcessLauncher",
            "device_isolation": false,
            "device_isolation_action": "error",
            "start_method": "spawn"
        },
        "environment": {
            "cpu": " AMD EPYC 7763 64-Core Processor",
            "cpu_count": 4,
            "cpu_ram_mb": 16757.354496,
            "system": "Linux",
            "machine": "x86_64",
            "platform": "Linux-6.5.0-1021-azure-x86_64-with-glibc2.35",
            "processor": "x86_64",
            "python_version": "3.10.14",
            "optimum_benchmark_version": "0.2.0",
            "optimum_benchmark_commit": "e65976e2695b67f37a76baa42b75347a3a733547",
            "transformers_version": "4.40.2",
            "transformers_commit": null,
            "accelerate_version": "0.30.1",
            "accelerate_commit": null,
            "diffusers_version": "0.27.2",
            "diffusers_commit": null,
            "optimum_version": null,
            "optimum_commit": null,
            "timm_version": "1.0.3",
            "timm_commit": null,
            "peft_version": null,
            "peft_commit": null
        }
    },
    "report": {
        "prefill": {
            "memory": {
                "unit": "MB",
                "max_ram": 955.957248,
                "max_global_vram": null,
                "max_process_vram": null,
                "max_reserved": null,
                "max_allocated": null
            },
            "latency": {
                "unit": "s",
                "count": 12,
                "total": 0.6236697900000365,
                "mean": 0.051972482500003046,
                "stdev": 0.0009896066504936835,
                "p50": 0.05210255250000273,
                "p90": 0.05304659979999542,
                "p95": 0.053136141500004896,
                "p99": 0.05319952350001245,
                "values": [
                    0.05155812900000001,
                    0.052824126999979626,
                    0.04986573400000793,
                    0.05278003500001205,
                    0.053071318999997175,
                    0.05040250700000115,
                    0.05203853800000502,
                    0.053215369000014334,
                    0.05216656700000044,
                    0.0519547510000109,
                    0.05239303100000825,
                    0.05139968299999964
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 38.481902418263026
            },
            "energy": {
                "unit": "kWh",
                "cpu": 1.8088019857502946e-06,
                "ram": 7.559147743902176e-08,
                "gpu": 0.0,
                "total": 1.8843934631893163e-06
            },
            "efficiency": {
                "unit": "tokens/kWh",
                "value": 1061349.4681811414
            }
        },
        "decode": {
            "memory": {
                "unit": "MB",
                "max_ram": 955.957248,
                "max_global_vram": null,
                "max_process_vram": null,
                "max_reserved": null,
                "max_allocated": null
            },
            "latency": {
                "unit": "s",
                "count": 12,
                "total": 0.39321803700005376,
                "mean": 0.03276816975000448,
                "stdev": 0.0006748811504500961,
                "p50": 0.0327170760000115,
                "p90": 0.03378614469999661,
                "p95": 0.03385203670000862,
                "p99": 0.03386303054002099,
                "values": [
                    0.03271283300000505,
                    0.032177411000020584,
                    0.033228828000005706,
                    0.03329431000000227,
                    0.03278884600001675,
                    0.03386577900002408,
                    0.03238653299999328,
                    0.032143327,
                    0.03146428799999512,
                    0.03384079299999598,
                    0.032593769999976985,
                    0.03272131900001796
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 30.517420033807753
            },
            "energy": {
                "unit": "kWh",
                "cpu": 1.0553131199846372e-06,
                "ram": 4.4104752275368847e-08,
                "gpu": 0.0,
                "total": 1.0994178722600061e-06
            },
            "efficiency": {
                "unit": "tokens/kWh",
                "value": 909572.261131576
            }
        },
        "per_token": {
            "memory": null,
            "latency": {
                "unit": "s",
                "count": 11,
                "total": 0.9328658499999847,
                "mean": 0.08480598636363497,
                "stdev": 0.0009291750153929013,
                "p50": 0.08503700500000377,
                "p90": 0.08592041700001118,
                "p95": 0.08601157249999858,
                "p99": 0.0860844968999885,
                "values": [
                    0.08507433500000161,
                    0.08307965399998807,
                    0.08610272799998597,
                    0.08592041700001118,
                    0.08426797600000668,
                    0.08445250200000487,
                    0.085390775999997,
                    0.0836498909999932,
                    0.08576265100001024,
                    0.08503700500000377,
                    0.08412791499998207
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 11.791620413589136
            },
            "energy": null,
            "efficiency": null
        }
    }
}