{
    "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": "a84403909b6e036e6ac3ccba23f1df81c197a364",
            "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": 945.291264,
                "max_global_vram": null,
                "max_process_vram": null,
                "max_reserved": null,
                "max_allocated": null
            },
            "latency": {
                "unit": "s",
                "count": 13,
                "total": 0.6234438079999904,
                "mean": 0.04795721599999926,
                "stdev": 0.0010602164960515198,
                "p50": 0.047924813000008726,
                "p90": 0.04920104799999194,
                "p95": 0.04958000759999095,
                "p99": 0.05001942791998999,
                "values": [
                    0.04716649900001357,
                    0.047851404999988745,
                    0.04652067500001067,
                    0.04914994399999273,
                    0.047924813000008726,
                    0.04620889000000261,
                    0.047777336000024206,
                    0.048379627000002756,
                    0.049213823999991746,
                    0.050129282999989755,
                    0.04814937399999053,
                    0.047015014999999494,
                    0.0479571229999749
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 41.70383868821807
            },
            "energy": {
                "unit": "kWh",
                "cpu": 1.5945832994249135e-06,
                "ram": 6.664023496911113e-08,
                "gpu": 0.0,
                "total": 1.6612235343940246e-06
            },
            "efficiency": {
                "unit": "tokens/kWh",
                "value": 1203931.896335404
            }
        },
        "decode": {
            "memory": {
                "unit": "MB",
                "max_ram": 945.291264,
                "max_global_vram": null,
                "max_process_vram": null,
                "max_reserved": null,
                "max_allocated": null
            },
            "latency": {
                "unit": "s",
                "count": 13,
                "total": 0.3903413210000224,
                "mean": 0.030026255461540183,
                "stdev": 0.00042299475511682947,
                "p50": 0.03017872099999863,
                "p90": 0.030508497000005262,
                "p95": 0.03058195120000846,
                "p99": 0.030634889440009375,
                "values": [
                    0.02943988399999853,
                    0.0305378360000077,
                    0.030256769000004624,
                    0.0300096849999818,
                    0.029974068000001353,
                    0.030391140999995514,
                    0.030334002999978793,
                    0.029495048000001134,
                    0.03017872099999863,
                    0.030648124000009602,
                    0.03019704600001205,
                    0.0293241660000092,
                    0.029554830000023458
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 33.30418610741766
            },
            "energy": {
                "unit": "kWh",
                "cpu": 9.459344645341237e-07,
                "ram": 3.953454741948834e-08,
                "gpu": 0.0,
                "total": 9.854690119536124e-07
            },
            "efficiency": {
                "unit": "tokens/kWh",
                "value": 1014745.251114067
            }
        },
        "per_token": {
            "memory": null,
            "latency": {
                "unit": "s",
                "count": 12,
                "total": 0.9373783519999961,
                "mean": 0.07811486266666634,
                "stdev": 0.001230883767789822,
                "p50": 0.07804928250000387,
                "p90": 0.07936456369999974,
                "p95": 0.08001338909999874,
                "p99": 0.08063352421999553,
                "values": [
                    0.0784812240000008,
                    0.0767301150000037,
                    0.07923319699997933,
                    0.07789698600001316,
                    0.0765836610000008,
                    0.07817973699999925,
                    0.07791882800000849,
                    0.079379160000002,
                    0.08078855799999474,
                    0.07836057699998378,
                    0.07630155100000025,
                    0.0775247580000098
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 12.801661116236286
            },
            "energy": null,
            "efficiency": null
        }
    }
}