{
    "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",
            "numactl": false,
            "numactl_kwargs": {},
            "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": "0.2.1",
            "optimum_benchmark_commit": "619f91dd928d405e72e9e7724b958f1538ca7d80",
            "transformers_version": "4.41.0",
            "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": 945.229824,
                "max_global_vram": null,
                "max_process_vram": null,
                "max_reserved": null,
                "max_allocated": null
            },
            "latency": {
                "unit": "s",
                "count": 13,
                "total": 0.6428320870000448,
                "mean": 0.049448622076926525,
                "stdev": 0.0020078466915931406,
                "p50": 0.04985194600001819,
                "p90": 0.05064141419998123,
                "p95": 0.050848530999991225,
                "p99": 0.051088823800013186,
                "values": [
                    0.05031153199999494,
                    0.05114889700001868,
                    0.049722705999982963,
                    0.05061392300001444,
                    0.048283080000004475,
                    0.04985194600001819,
                    0.050306051000006846,
                    0.04910588000001326,
                    0.050463232999987895,
                    0.04978999100001147,
                    0.050648286999972925,
                    0.049645063000014034,
                    0.042941498000004685
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 40.446020859562644
            },
            "energy": {
                "unit": "kWh",
                "cpu": 1.6938906360939505e-06,
                "ram": 7.0788632798761e-08,
                "gpu": 0.0,
                "total": 1.7646792688927113e-06
            },
            "efficiency": {
                "unit": "tokens/kWh",
                "value": 1133350.4253466672
            }
        },
        "decode": {
            "memory": {
                "unit": "MB",
                "max_ram": 945.229824,
                "max_global_vram": null,
                "max_process_vram": null,
                "max_reserved": null,
                "max_allocated": null
            },
            "latency": {
                "unit": "s",
                "count": 13,
                "total": 0.3983753859999126,
                "mean": 0.03064426046153174,
                "stdev": 0.0014700787956508639,
                "p50": 0.03131875899998704,
                "p90": 0.031569557000005945,
                "p95": 0.03166596420000474,
                "p99": 0.031768659239995715,
                "values": [
                    0.03179433299999346,
                    0.02979006799998274,
                    0.031436087000003,
                    0.031399288999978126,
                    0.03131875899998704,
                    0.03143513599999892,
                    0.03112569999998982,
                    0.03079246999999441,
                    0.03158038500001226,
                    0.03152624499998069,
                    0.031199818000004598,
                    0.02839166699999396,
                    0.02658542899999361
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 32.63253819602914
            },
            "energy": {
                "unit": "kWh",
                "cpu": 1.003325945321297e-06,
                "ram": 4.193258926862206e-08,
                "gpu": 0.0,
                "total": 1.0452585345899192e-06
            },
            "efficiency": {
                "unit": "tokens/kWh",
                "value": 956701.1097329378
            }
        },
        "per_token": {
            "memory": null,
            "latency": {
                "unit": "s",
                "count": 12,
                "total": 0.9594809250000083,
                "mean": 0.07995674375000068,
                "stdev": 0.003323258659392264,
                "p50": 0.08127134800000135,
                "p90": 0.08198370220000015,
                "p95": 0.08202318359999908,
                "p99": 0.08204872471999891,
                "values": [
                    0.08096021399998676,
                    0.08120404800001779,
                    0.08199706199999923,
                    0.079628649,
                    0.08134090199999378,
                    0.08145156800000564,
                    0.07989444300000059,
                    0.08205510999999888,
                    0.08133864799998491,
                    0.08186346400000843,
                    0.07820807799998875,
                    0.0695387390000235
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 12.506762445537827
            },
            "energy": null,
            "efficiency": null
        }
    }
}