IlyasMoutawwakil's picture
Upload cpu_inference_transformers_text-generation_openai-community/gpt2/benchmark.json with huggingface_hub
9419563 verified
raw
history blame
8.09 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": 16757.354496,
"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": "c78698d47f8aff03fc8162bfacd2946f5882e025",
"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": 956.997632,
"max_global_vram": null,
"max_process_vram": null,
"max_reserved": null,
"max_allocated": null
},
"latency": {
"unit": "s",
"count": 12,
"total": 0.6255316360000052,
"mean": 0.05212763633333376,
"stdev": 0.0018987139371279534,
"p50": 0.05251503550000791,
"p90": 0.053433454099996426,
"p95": 0.054302879250002434,
"p99": 0.05512467825000727,
"values": [
0.052909083000002965,
0.052231663000014805,
0.052043943000001036,
0.05279840800000102,
0.053172913999986804,
0.05202980699999671,
0.05186763599999722,
0.053462402999997494,
0.0528243550000127,
0.05533012800000847,
0.04919050399999492,
0.047670791999991025
]
},
"throughput": {
"unit": "tokens/s",
"value": 38.367364044877505
},
"energy": {
"unit": "kWh",
"cpu": 1.8424941433800592e-06,
"ram": 7.699700820690613e-08,
"gpu": 0.0,
"total": 1.9194911515869657e-06
},
"efficiency": {
"unit": "tokens/kWh",
"value": 1041942.8077834443
}
},
"decode": {
"memory": {
"unit": "MB",
"max_ram": 956.997632,
"max_global_vram": null,
"max_process_vram": null,
"max_reserved": null,
"max_allocated": null
},
"latency": {
"unit": "s",
"count": 12,
"total": 0.386311921000015,
"mean": 0.03219266008333458,
"stdev": 0.0018401079157075837,
"p50": 0.03256909950000875,
"p90": 0.033920342100014975,
"p95": 0.034344610750002856,
"p99": 0.03471016934999056,
"values": [
0.03310820800001579,
0.03293685899998877,
0.03397074400001543,
0.032154352999981484,
0.03346672500001091,
0.032274086000001034,
0.032335600000010345,
0.03280259900000715,
0.03480155899998749,
0.03136097600000198,
0.028449919999985696,
0.028650292000008903
]
},
"throughput": {
"unit": "tokens/s",
"value": 31.062981356973282
},
"energy": {
"unit": "kWh",
"cpu": 1.1060721344417995e-06,
"ram": 4.6229628038968345e-08,
"gpu": 0.0,
"total": 1.1523017624807676e-06
},
"efficiency": {
"unit": "tokens/kWh",
"value": 867828.2309029188
}
},
"per_token": {
"memory": null,
"latency": {
"unit": "s",
"count": 11,
"total": 0.9262350150000032,
"mean": 0.08420318318181846,
"stdev": 0.0035513016434451003,
"p50": 0.08519879800002172,
"p90": 0.08692494000001716,
"p95": 0.08727531700000668,
"p99": 0.0875556185999983,
"values": [
0.08519879800002172,
0.08603374099999428,
0.08500937599998792,
0.08662094400000342,
0.08435044000000858,
0.08423834999999258,
0.08623108899999465,
0.0876256939999962,
0.08692494000001716,
0.07766662199998109,
0.07633502100000555
]
},
"throughput": {
"unit": "tokens/s",
"value": 11.876035586929266
},
"energy": null,
"efficiency": null
}
}
}