IlyasMoutawwakil's picture
Upload cpu_inference_transformers_text-generation_openai-community/gpt2/benchmark.json with huggingface_hub
16658ec verified
raw
history blame
9.8 kB
{
"config": {
"name": "cpu_inference_transformers_text-generation_openai-community/gpt2",
"backend": {
"name": "pytorch",
"version": "2.3.1+cpu",
"_target_": "optimum_benchmark.backends.pytorch.backend.PyTorchBackend",
"task": "text-generation",
"library": "transformers",
"model_type": "gpt2",
"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,
"memory": true,
"latency": 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": 16757.334016,
"system": "Linux",
"machine": "x86_64",
"platform": "Linux-6.5.0-1023-azure-x86_64-with-glibc2.35",
"processor": "x86_64",
"python_version": "3.10.14",
"optimum_benchmark_version": "0.3.1",
"optimum_benchmark_commit": "bc977b19adf337e8eb02985f3cceb35e1a783a1c",
"transformers_version": "4.42.4",
"transformers_commit": null,
"accelerate_version": "0.32.1",
"accelerate_commit": null,
"diffusers_version": "0.29.2",
"diffusers_commit": null,
"optimum_version": null,
"optimum_commit": null,
"timm_version": "1.0.7",
"timm_commit": null,
"peft_version": null,
"peft_commit": null
}
},
"report": {
"load": {
"memory": {
"unit": "MB",
"max_ram": 1111.105536,
"max_global_vram": null,
"max_process_vram": null,
"max_reserved": null,
"max_allocated": null
},
"latency": {
"unit": "s",
"count": 1,
"total": 4.7929352140000105,
"mean": 4.7929352140000105,
"stdev": 0.0,
"p50": 4.7929352140000105,
"p90": 4.7929352140000105,
"p95": 4.7929352140000105,
"p99": 4.7929352140000105,
"values": [
4.7929352140000105
]
},
"throughput": null,
"energy": {
"unit": "kWh",
"cpu": 6.834825807147557e-05,
"ram": 2.856749149019417e-06,
"gpu": 0,
"total": 7.120500722049499e-05
},
"efficiency": null
},
"prefill": {
"memory": {
"unit": "MB",
"max_ram": 964.050944,
"max_global_vram": null,
"max_process_vram": null,
"max_reserved": null,
"max_allocated": null
},
"latency": {
"unit": "s",
"count": 16,
"total": 0.6610148060000256,
"mean": 0.0413134253750016,
"stdev": 0.0009603430324635616,
"p50": 0.041168685999991794,
"p90": 0.04233700399998952,
"p95": 0.04320748450001588,
"p99": 0.043810029700028966,
"values": [
0.04171758399996861,
0.04168280899995125,
0.04099478000000545,
0.041031760000009854,
0.04117999699997199,
0.04153460200001291,
0.04295642400001043,
0.04042434199999434,
0.04140254499998264,
0.04131998000002568,
0.0411573750000116,
0.043960666000032234,
0.04092735400001857,
0.040352868000013586,
0.04000683999998955,
0.040364880000026915
]
},
"throughput": {
"unit": "tokens/s",
"value": 48.41041336674501
},
"energy": {
"unit": "kWh",
"cpu": 1.5958502027723528e-06,
"ram": 6.669236981249469e-08,
"gpu": 0.0,
"total": 1.6625425725848473e-06
},
"efficiency": {
"unit": "tokens/kWh",
"value": 1202976.7134867944
}
},
"decode": {
"memory": {
"unit": "MB",
"max_ram": 964.968448,
"max_global_vram": null,
"max_process_vram": null,
"max_reserved": null,
"max_allocated": null
},
"latency": {
"unit": "s",
"count": 16,
"total": 0.4018238840000663,
"mean": 0.025113992750004144,
"stdev": 0.000421922355316923,
"p50": 0.02496320550000064,
"p90": 0.025659635000010894,
"p95": 0.025876258000025132,
"p99": 0.026040744400026482,
"values": [
0.02580772200002457,
0.02608186600002682,
0.024947521000001416,
0.0248139199999855,
0.02469160200001852,
0.02466596300001811,
0.025484737999988738,
0.024966195999979846,
0.02456152799999245,
0.02495480499999303,
0.02543860800000175,
0.025177202000008947,
0.024960215000021435,
0.024733359999970617,
0.025027090000037333,
0.02551154799999722
]
},
"throughput": {
"unit": "tokens/s",
"value": 39.8184394633878
},
"energy": {
"unit": "kWh",
"cpu": 9.550531738334227e-07,
"ram": 3.991597315357703e-08,
"gpu": 0.0,
"total": 9.949691469870006e-07
},
"efficiency": {
"unit": "tokens/kWh",
"value": 1005056.2904671306
}
},
"per_token": {
"memory": null,
"latency": {
"unit": "s",
"count": 16,
"total": 0.39643982500001584,
"mean": 0.02477748906250099,
"stdev": 0.00042147732616428066,
"p50": 0.024652052500016453,
"p90": 0.025349012999981824,
"p95": 0.025548912250016542,
"p99": 0.0257170232500215,
"values": [
0.025478866000014477,
0.025759051000022737,
0.024626729999965846,
0.0244951430000242,
0.024349130000018704,
0.02430610900000829,
0.02515880699996842,
0.024595862000012403,
0.024204880000013418,
0.02462709100001348,
0.024931728000012754,
0.0248595669999645,
0.024677014000019426,
0.024447974999986855,
0.024702712000021165,
0.02521915999994917
]
},
"throughput": {
"unit": "tokens/s",
"value": 40.35921466769732
},
"energy": null,
"efficiency": null
}
}
}