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": "6fcf23d15753ee501a4776b6dfde05d914257752",
            "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": 956.37504,
                "max_global_vram": null,
                "max_process_vram": null,
                "max_reserved": null,
                "max_allocated": null
            },
            "latency": {
                "unit": "s",
                "count": 11,
                "total": 0.6311531610000145,
                "mean": 0.05737756009091041,
                "stdev": 0.0013297514938054363,
                "p50": 0.05802002000001494,
                "p90": 0.05829856899998731,
                "p95": 0.05893202800000097,
                "p99": 0.0594387952000119,
                "values": [
                    0.056171938999995064,
                    0.05802002000001494,
                    0.054376826999998684,
                    0.05822428000001878,
                    0.05803257300001974,
                    0.056439938000011125,
                    0.058028574999980265,
                    0.05829856899998731,
                    0.05667590799998834,
                    0.057319044999985636,
                    0.05956548700001463
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 34.856832476514356
            },
            "energy": {
                "unit": "kWh",
                "cpu": 1.9997935824924045e-06,
                "ram": 8.357228617506962e-08,
                "gpu": 0.0,
                "total": 2.0833658686674743e-06
            },
            "efficiency": {
                "unit": "tokens/kWh",
                "value": 959985.007952158
            }
        },
        "decode": {
            "memory": {
                "unit": "MB",
                "max_ram": 956.37504,
                "max_global_vram": null,
                "max_process_vram": null,
                "max_reserved": null,
                "max_allocated": null
            },
            "latency": {
                "unit": "s",
                "count": 11,
                "total": 0.397706138000018,
                "mean": 0.03615510345454709,
                "stdev": 0.0008480934039526884,
                "p50": 0.03601686500002188,
                "p90": 0.037410609999994904,
                "p95": 0.037456891499999756,
                "p99": 0.03749391670000364,
                "values": [
                    0.03464203600000815,
                    0.035462422999984256,
                    0.035859702999999854,
                    0.0358088280000004,
                    0.03538275500000054,
                    0.03699494599999298,
                    0.03750317300000461,
                    0.03601956900001824,
                    0.037410609999994904,
                    0.03601686500002188,
                    0.03660522999999216
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 27.6586126010444
            },
            "energy": {
                "unit": "kWh",
                "cpu": 1.335924731360542e-06,
                "ram": 5.582244644143941e-08,
                "gpu": 0.0,
                "total": 1.3917471778019815e-06
            },
            "efficiency": {
                "unit": "tokens/kWh",
                "value": 718521.3061321405
            }
        },
        "per_token": {
            "memory": null,
            "latency": {
                "unit": "s",
                "count": 10,
                "total": 0.9382666349999909,
                "mean": 0.09382666349999909,
                "stdev": 0.0015058675468307941,
                "p50": 0.09380351550001365,
                "p90": 0.09562797209999019,
                "p95": 0.09592090004998255,
                "p99": 0.09615524240997644,
                "values": [
                    0.09351868000001673,
                    0.09024152899999649,
                    0.09409728699998254,
                    0.09343966299999806,
                    0.09340764300000615,
                    0.09556287699999189,
                    0.09437151700001323,
                    0.09408835100001056,
                    0.0933252600000003,
                    0.09621382799997491
                ]
            },
            "throughput": {
                "unit": "tokens/s",
                "value": 10.657951191028014
            },
            "energy": null,
            "efficiency": null
        }
    }
}