{
    "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": 16757.3504,
            "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": "b50afdd89f3c69090dee699e5e398c44916e2d64",
            "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": 956.7232,
                "max_global_vram": null,
                "max_process_vram": null,
                "max_reserved": null,
                "max_allocated": null
            },
            "latency": {
                "unit": "s",
                "count": 15,
                "total": 0.6569378359998268,
                "mean": 0.04379585573332179,
                "stdev": 0.0011521062606398742,
                "p50": 0.0433752600000048,
                "p90": 0.04505936279999787,
                "p95": 0.04570637779999629,
                "p99": 0.04675128515998381,
                "values": [
                    0.045146606000002976,
                    0.0433752600000048,
                    0.0430425170000035,
                    0.04492849799999021,
                    0.04305154299999003,
                    0.043888229000003776,
                    0.04340799099998094,
                    0.044274111999982324,
                    0.0427798939999775,
                    0.043185523999966335,
                    0.044431446999965374,
                    0.042782579999993686,
                    0.04289457800001628,
                    0.04701251199998069,
                    0.04273654499996837
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 45.66642132027833
            },
            "energy": {
                "unit": "kWh",
                "cpu": 1.6711255466496504e-06,
                "ram": 6.983898169840883e-08,
                "gpu": 0.0,
                "total": 1.7409645283480592e-06
            },
            "efficiency": {
                "unit": "tokens/kWh",
                "value": 1148788.4832999615
            }
        },
        "decode": {
            "memory": {
                "unit": "MB",
                "max_ram": 956.7232,
                "max_global_vram": null,
                "max_process_vram": null,
                "max_reserved": null,
                "max_allocated": null
            },
            "latency": {
                "unit": "s",
                "count": 15,
                "total": 0.40627674100016975,
                "mean": 0.027085116066677982,
                "stdev": 0.0005490786431801712,
                "p50": 0.026920791999998528,
                "p90": 0.027575365199982115,
                "p95": 0.02803870459998734,
                "p99": 0.028649240120005286,
                "values": [
                    0.026981576000025598,
                    0.027711631999977726,
                    0.028801874000009775,
                    0.02723999000005506,
                    0.02690345100000968,
                    0.02697736900000791,
                    0.0273709649999887,
                    0.026432368000030237,
                    0.027107363000027362,
                    0.02664768200003209,
                    0.026920791999998528,
                    0.026917747000027248,
                    0.02663911399997687,
                    0.02683371799997758,
                    0.02679110000002538
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 36.920646658415855
            },
            "energy": {
                "unit": "kWh",
                "cpu": 1.0024787651167968e-06,
                "ram": 4.189791105792847e-08,
                "gpu": 0.0,
                "total": 1.0443766761747255e-06
            },
            "efficiency": {
                "unit": "tokens/kWh",
                "value": 957508.9360121815
            }
        },
        "per_token": {
            "memory": null,
            "latency": {
                "unit": "s",
                "count": 14,
                "total": 0.9912976520000143,
                "mean": 0.07080697514285816,
                "stdev": 0.001185979296676912,
                "p50": 0.07075481100000047,
                "p90": 0.07215794469998968,
                "p95": 0.07289375664999795,
                "p99": 0.07366563373001497,
                "values": [
                    0.07109474700001783,
                    0.07165329299999712,
                    0.0723742239999865,
                    0.0699773549999918,
                    0.07089456300002439,
                    0.07073776900000439,
                    0.07077185299999655,
                    0.06989950000001954,
                    0.06984415599998783,
                    0.07137460199999168,
                    0.06972029399997837,
                    0.06954488299999184,
                    0.07385860300001923,
                    0.06955181000000721
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 14.12290241155519
            },
            "energy": null,
            "efficiency": null
        }
    }
}