{
    "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": "1981150adc21fc49153cb832dbc6f0ecca02bc2a",
            "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": 946.352128,
                "max_global_vram": null,
                "max_process_vram": null,
                "max_reserved": null,
                "max_allocated": null
            },
            "latency": {
                "unit": "s",
                "count": 13,
                "total": 0.6274317090000352,
                "mean": 0.048263977615387324,
                "stdev": 0.0033652210581413544,
                "p50": 0.05000621199999955,
                "p90": 0.051905234600008045,
                "p95": 0.052314373200010776,
                "p99": 0.05242016904000593,
                "values": [
                    0.050053219999995235,
                    0.05062133299998095,
                    0.05041648400001009,
                    0.05058210100000338,
                    0.04833408099997882,
                    0.05000621199999955,
                    0.05244661800000472,
                    0.04748237200001881,
                    0.0466254529999901,
                    0.05222621000001482,
                    0.044000795000016524,
                    0.04215734600001042,
                    0.04247948400001178
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 41.43877274139892
            },
            "energy": {
                "unit": "kWh",
                "cpu": 1.683818300565084e-06,
                "ram": 7.03694987358716e-08,
                "gpu": 0.0,
                "total": 1.7541877993009554e-06
            },
            "efficiency": {
                "unit": "tokens/kWh",
                "value": 1140128.7825607958
            }
        },
        "decode": {
            "memory": {
                "unit": "MB",
                "max_ram": 946.352128,
                "max_global_vram": null,
                "max_process_vram": null,
                "max_reserved": null,
                "max_allocated": null
            },
            "latency": {
                "unit": "s",
                "count": 13,
                "total": 0.38780337700001155,
                "mean": 0.02983102900000089,
                "stdev": 0.002023949214410245,
                "p50": 0.03038238799999249,
                "p90": 0.03163492020000262,
                "p95": 0.03175620740000795,
                "p99": 0.03187531508001484,
                "values": [
                    0.0311774630000059,
                    0.029199142999999594,
                    0.0314423330000011,
                    0.03165695100000221,
                    0.031905092000016566,
                    0.03154679700000429,
                    0.03147077599999193,
                    0.03016786000000593,
                    0.029543842999999015,
                    0.03038238799999249,
                    0.02623006700000019,
                    0.02670120999999881,
                    0.02637945399999353
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 33.52214233038928
            },
            "energy": {
                "unit": "kWh",
                "cpu": 1.0219877075265956e-06,
                "ram": 4.271205439965324e-08,
                "gpu": 0.0,
                "total": 1.0646997619262488e-06
            },
            "efficiency": {
                "unit": "tokens/kWh",
                "value": 939231.9184807609
            }
        },
        "per_token": {
            "memory": null,
            "latency": {
                "unit": "s",
                "count": 12,
                "total": 0.9344495689999803,
                "mean": 0.07787079741666503,
                "stdev": 0.005296569182113948,
                "p50": 0.08007116699999983,
                "p90": 0.08252702070002157,
                "p95": 0.08317721535000544,
                "p99": 0.08378542866998658,
                "values": [
                    0.07991126499999268,
                    0.08187582799999404,
                    0.08227359599999318,
                    0.08023106900000698,
                    0.08155465300001197,
                    0.08393748199998186,
                    0.07765196500000116,
                    0.07622493799999575,
                    0.08255517900002474,
                    0.0704801739999823,
                    0.06886468700000137,
                    0.06888873299999432
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 12.841784509400584
            },
            "energy": null,
            "efficiency": null
        }
    }
}