IlyasMoutawwakil's picture
Upload cuda_inference_transformers_text-generation_openai-community/gpt2/benchmark.json with huggingface_hub
ffe090d verified
raw
history blame
12.8 kB
{
"config": {
"name": "cuda_inference_transformers_text-generation_openai-community/gpt2",
"backend": {
"name": "pytorch",
"version": "2.2.2+rocm5.7",
"_target_": "optimum_benchmark.backends.pytorch.backend.PyTorchBackend",
"task": "text-generation",
"library": "transformers",
"model": "openai-community/gpt2",
"processor": "openai-community/gpt2",
"device": "cuda",
"device_ids": "0",
"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": false,
"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": true,
"device_isolation_action": "error",
"numactl": false,
"numactl_kwargs": {},
"start_method": "spawn"
},
"environment": {
"cpu": " AMD EPYC 7763 64-Core Processor",
"cpu_count": 128,
"cpu_ram_mb": 1082015.236096,
"system": "Linux",
"machine": "x86_64",
"platform": "Linux-5.15.0-84-generic-x86_64-with-glibc2.35",
"processor": "x86_64",
"python_version": "3.10.12",
"gpu": [
"Advanced Micro Devices, Inc. [AMD/ATI]"
],
"gpu_count": 1,
"gpu_vram_mb": 68702699520,
"optimum_benchmark_version": "0.2.1",
"optimum_benchmark_commit": "dc29eecfd7850e87154a5a92442eb64d237d0070",
"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": 1047.138304,
"max_global_vram": 2087.612416,
"max_process_vram": 270334.79168,
"max_reserved": 725.614592,
"max_allocated": 512.83456
},
"latency": {
"unit": "s",
"count": 52,
"total": 0.4912512798309325,
"mean": 0.009447139996748705,
"stdev": 0.0002835958066960971,
"p50": 0.009379683971405029,
"p90": 0.00962755699157715,
"p95": 0.009759940814971923,
"p99": 0.010605658102035525,
"values": [
0.01112928581237793,
0.009813445091247559,
0.009525124549865722,
0.009408644676208496,
0.009344803810119628,
0.009319205284118653,
0.009465124130249023,
0.009389445304870606,
0.009438244819641114,
0.009535204887390137,
0.009320804595947266,
0.009321603775024414,
0.010102564811706544,
0.00936624526977539,
0.009604804992675782,
0.009380643844604492,
0.009536805152893067,
0.009426883697509766,
0.009397125244140625,
0.00953856372833252,
0.009716164588928223,
0.009630084991455078,
0.009485604286193848,
0.009452964782714844,
0.009378724098205567,
0.00935648536682129,
0.009403843879699707,
0.009335044860839845,
0.009694244384765625,
0.009505764961242676,
0.009347204208374024,
0.00935856533050537,
0.009328483581542969,
0.009319845199584962,
0.009411844253540039,
0.009348164558410645,
0.00938944435119629,
0.00936368465423584,
0.0093812837600708,
0.009319364547729492,
0.009276324272155762,
0.009304644584655762,
0.00925952434539795,
0.00924832534790039,
0.00927568531036377,
0.009255364418029785,
0.009439205169677735,
0.009254083633422852,
0.009292964935302734,
0.009261444091796875,
0.009247044563293456,
0.009245285034179688
]
},
"throughput": {
"unit": "tokens/s",
"value": 211.7042830622086
},
"energy": null,
"efficiency": null
},
"decode": {
"memory": {
"unit": "MB",
"max_ram": 1047.138304,
"max_global_vram": 2087.612416,
"max_process_vram": 269365.850112,
"max_reserved": 725.614592,
"max_allocated": 513.058304
},
"latency": {
"unit": "s",
"count": 52,
"total": 0.5174629678726195,
"mean": 0.0099512109206273,
"stdev": 0.00017757253935113287,
"p50": 0.009924084663391113,
"p90": 0.01012277317047119,
"p95": 0.010153604888916016,
"p99": 0.010564165010452271,
"values": [
0.01097216510772705,
0.009970085144042969,
0.00978128433227539,
0.009754084587097167,
0.009810725212097167,
0.009878725051879882,
0.009826404571533203,
0.00984144401550293,
0.009904484748840332,
0.009760964393615722,
0.0098931245803833,
0.009843364715576172,
0.009853764533996582,
0.009830563545227052,
0.009884804725646972,
0.009896644592285156,
0.009874244689941407,
0.009737924575805664,
0.009849445343017578,
0.009949284553527832,
0.010144965171813965,
0.010025284767150878,
0.009964645385742188,
0.009915844917297363,
0.009809764862060547,
0.010128805160522461,
0.009918564796447754,
0.009891683578491212,
0.010138884544372558,
0.010068485260009766,
0.010025924682617187,
0.00996688461303711,
0.009929604530334472,
0.01000384521484375,
0.009973605155944825,
0.010164164543151855,
0.010054405212402345,
0.009987845420837402,
0.010040644645690919,
0.009940645217895509,
0.009906885147094726,
0.010172164916992188,
0.00987840461730957,
0.009852964401245118,
0.009888324737548828,
0.00977680492401123,
0.009959525108337402,
0.009942564964294433,
0.010014083862304688,
0.009931685447692871,
0.00995168399810791,
0.00997984504699707
]
},
"throughput": {
"unit": "tokens/s",
"value": 100.49028283855955
},
"energy": null,
"efficiency": null
},
"per_token": {
"memory": null,
"latency": {
"unit": "s",
"count": 51,
"total": 0.9889406089782717,
"mean": 0.019390992332907284,
"stdev": 0.00019725036095099918,
"p50": 0.01936465072631836,
"p90": 0.01959185028076172,
"p95": 0.01984680938720703,
"p99": 0.019995049476623533,
"values": [
0.019774248123168945,
0.01940401077270508,
0.019158727645874025,
0.019157289505004883,
0.01927872848510742,
0.01936465072631836,
0.01921264839172363,
0.01947104835510254,
0.01931265068054199,
0.01915840721130371,
0.019230730056762697,
0.02006576919555664,
0.019250728607177733,
0.019483367919921876,
0.019430089950561523,
0.019366567611694335,
0.019273929595947266,
0.019322727203369142,
0.019482570648193358,
0.019919370651245116,
0.0197100887298584,
0.01952640914916992,
0.01941200828552246,
0.019246889114379884,
0.019421287536621094,
0.0193748893737793,
0.01926144790649414,
0.01992432975769043,
0.019518407821655273,
0.01945969009399414,
0.019310407638549805,
0.019365129470825197,
0.01930400848388672,
0.019461929321289064,
0.019533447265625,
0.01946320915222168,
0.019401927947998047,
0.01959185028076172,
0.019257768630981446,
0.01924272918701172,
0.019318567276000977,
0.019391210556030273,
0.0191534481048584,
0.019209287643432618,
0.01914208984375,
0.019285127639770507,
0.019284170150756837,
0.019365608215332033,
0.019297927856445313,
0.01926272773742676,
0.019324327468872072
]
},
"throughput": {
"unit": "tokens/s",
"value": 51.570336516659864
},
"energy": null,
"efficiency": null
}
}
}