IlyasMoutawwakil's picture
Upload cpu_inference_transformers_text-generation_openai-community/gpt2/benchmark.json with huggingface_hub
8da8974 verified
raw
history blame
8.21 kB
{
"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": 16757.3504,
"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": "93f16ddd3ee4bf4a82d1c5d47e103207299d4a25",
"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": 955.61728,
"max_global_vram": null,
"max_process_vram": null,
"max_reserved": null,
"max_allocated": null
},
"latency": {
"unit": "s",
"count": 13,
"total": 0.6278269489999388,
"mean": 0.04829438069230299,
"stdev": 0.0008603489852715406,
"p50": 0.04815443199998981,
"p90": 0.049462257399989086,
"p95": 0.04975024719998942,
"p99": 0.04989901743999667,
"values": [
0.048497101999998904,
0.04993620999999848,
0.048109356999987085,
0.04880619900001193,
0.04815443199998981,
0.0470793629999946,
0.048038133999995125,
0.04962627199998337,
0.04762457100000006,
0.04877326799999082,
0.04850096900000267,
0.046743055000007416,
0.04793801699997857
]
},
"throughput": {
"unit": "tokens/s",
"value": 41.41268552013452
},
"energy": {
"unit": "kWh",
"cpu": 1.6732901334762574e-06,
"ram": 6.992635130230617e-08,
"gpu": 0.0,
"total": 1.7432164847785636e-06
},
"efficiency": {
"unit": "tokens/kWh",
"value": 1147304.4326184506
}
},
"decode": {
"memory": {
"unit": "MB",
"max_ram": 955.61728,
"max_global_vram": null,
"max_process_vram": null,
"max_reserved": null,
"max_allocated": null
},
"latency": {
"unit": "s",
"count": 13,
"total": 0.39210055000009447,
"mean": 0.030161580769238035,
"stdev": 0.0004024381716264546,
"p50": 0.03024565500001586,
"p90": 0.030713675400016884,
"p95": 0.030744500800017248,
"p99": 0.030755517760014756,
"values": [
0.03075827200001413,
0.030627097000007097,
0.03073532000001933,
0.03012799499998664,
0.029650433000000476,
0.030304475000008324,
0.02979204700000082,
0.03024565500001586,
0.029866916999992554,
0.030349568000019644,
0.030401355999998714,
0.02969920399999637,
0.029542211000034513
]
},
"throughput": {
"unit": "tokens/s",
"value": 33.15476094077621
},
"energy": {
"unit": "kWh",
"cpu": 9.92700766634058e-07,
"ram": 4.149310394308732e-08,
"gpu": 0.0,
"total": 1.034193870577146e-06
},
"efficiency": {
"unit": "tokens/kWh",
"value": 966936.6919008486
}
},
"per_token": {
"memory": null,
"latency": {
"unit": "s",
"count": 12,
"total": 0.9408672080000144,
"mean": 0.07840560066666787,
"stdev": 0.001158205158507909,
"p50": 0.07834177750000038,
"p90": 0.07989987149999536,
"p95": 0.08027617835000314,
"p99": 0.08056225007001075,
"values": [
0.08063376800001265,
0.07881117499999846,
0.07897489999999152,
0.0778723800000023,
0.07739306499999543,
0.07779285200001596,
0.07998360499999535,
0.07745219500000644,
0.07914626999999541,
0.07891870500000664,
0.07645614499998032,
0.07743214800001397
]
},
"throughput": {
"unit": "tokens/s",
"value": 12.754190918725076
},
"energy": null,
"efficiency": null
}
}
}