IlyasMoutawwakil's picture
Upload cuda_inference_transformers_text-generation_openai-community/gpt2/benchmark.json with huggingface_hub
a9ef565 verified
raw
history blame
13.3 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",
"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": "48414f58841d7ba7c7fd42d74fd524d1d23c3081",
"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": 1050.017792,
"max_global_vram": 2098.786304,
"max_process_vram": 291094.38464,
"max_reserved": 725.614592,
"max_allocated": 512.832512
},
"latency": {
"unit": "s",
"count": 56,
"total": 0.49914637088775643,
"mean": 0.008913328051567079,
"stdev": 0.0007680013742947539,
"p50": 0.008987133502960206,
"p90": 0.009295454502105713,
"p95": 0.009781295299530028,
"p99": 0.012086122560501103,
"values": [
0.011444817543029786,
0.00933393383026123,
0.009755695343017578,
0.009858095169067382,
0.009438413619995117,
0.00905073356628418,
0.008525612831115723,
0.008971053123474121,
0.009041133880615235,
0.009035054206848145,
0.009025774002075194,
0.012869939804077149,
0.009256975173950195,
0.009216013908386231,
0.009087214469909668,
0.00900321388244629,
0.009046573638916016,
0.00909105396270752,
0.009077134132385254,
0.009072813987731933,
0.009060334205627442,
0.0090833740234375,
0.009083053588867188,
0.00906977367401123,
0.009046414375305175,
0.009074893951416016,
0.009086414337158203,
0.008856013298034668,
0.008761453628540038,
0.00847713279724121,
0.008597773551940918,
0.008599213600158691,
0.008505772590637208,
0.009056014060974121,
0.008836013793945312,
0.008677613258361817,
0.008570094108581543,
0.008679533958435058,
0.008546572685241699,
0.008747533798217773,
0.008035853385925293,
0.00804977321624756,
0.008048172950744628,
0.008265453338623047,
0.009072493553161622,
0.0085529727935791,
0.009108334541320802,
0.00851905345916748,
0.008578573226928711,
0.008547053337097169,
0.008470093727111817,
0.008194091796875,
0.00802001190185547,
0.008012652397155762,
0.008003691673278808,
0.008051852226257324
]
},
"throughput": {
"unit": "tokens/s",
"value": 224.38307985852424
},
"energy": null,
"efficiency": null
},
"decode": {
"memory": {
"unit": "MB",
"max_ram": 1050.017792,
"max_global_vram": 2098.774016,
"max_process_vram": 285213.659136,
"max_reserved": 725.614592,
"max_allocated": 513.056256
},
"latency": {
"unit": "s",
"count": 56,
"total": 0.5011466875076297,
"mean": 0.008949047991207668,
"stdev": 0.00044721502938795327,
"p50": 0.009121774673461914,
"p90": 0.009487134456634521,
"p95": 0.009652574539184569,
"p99": 0.009789871406555175,
"values": [
0.00979233455657959,
0.009787856101989746,
0.009606413841247559,
0.009633614540100097,
0.00948001480102539,
0.00889841365814209,
0.008834094047546386,
0.009132015228271484,
0.009077774047851563,
0.009111534118652344,
0.009494254112243652,
0.009709454536437989,
0.009328813552856445,
0.009204174995422363,
0.009200974464416504,
0.00916081428527832,
0.009163373947143555,
0.009198094367980957,
0.009197134971618653,
0.009180653572082519,
0.009230414390563964,
0.009200814247131348,
0.009195694923400879,
0.009153293609619141,
0.009170093536376954,
0.009185613632202149,
0.009393614768981933,
0.008919854164123536,
0.00891201400756836,
0.00872721290588379,
0.008931532859802246,
0.008588652610778808,
0.009143854141235351,
0.009157294273376465,
0.008800013542175293,
0.008694092750549316,
0.008648492813110352,
0.008626572608947754,
0.008576172828674317,
0.00858193302154541,
0.008216172218322755,
0.00812353229522705,
0.008126412391662598,
0.009222093582153321,
0.008799054145812989,
0.009231213569641114,
0.009171053886413575,
0.00856449317932129,
0.008673773765563966,
0.008673453330993652,
0.00850705337524414,
0.00813009262084961,
0.008117612838745117,
0.008143532752990722,
0.008264171600341796,
0.008153932571411133
]
},
"throughput": {
"unit": "tokens/s",
"value": 111.7437297221434
},
"energy": null,
"efficiency": null
},
"per_token": {
"memory": null,
"latency": {
"unit": "s",
"count": 55,
"total": 0.9810565567016603,
"mean": 0.017837391940030183,
"stdev": 0.0010616067345246415,
"p50": 0.017969467163085937,
"p90": 0.018620060348510742,
"p95": 0.01913817272186279,
"p99": 0.02100902153015137,
"values": [
0.019028989791870118,
0.019375389099121093,
0.01955666923522949,
0.019036508560180663,
0.018104667663574218,
0.017431068420410156,
0.018079708099365235,
0.018211227416992187,
0.018163707733154295,
0.018494428634643553,
0.022713956832885743,
0.018665307998657225,
0.018453468322753908,
0.018309627532958984,
0.018207548141479492,
0.018240667343139647,
0.018325788497924805,
0.018304988861083983,
0.018285627365112304,
0.018324188232421875,
0.018322107315063476,
0.018307867050170897,
0.018256187438964844,
0.018248987197875977,
0.018292028427124022,
0.018552188873291017,
0.017754108428955077,
0.01772018814086914,
0.017275867462158204,
0.01752082824707031,
0.01724802589416504,
0.017643388748168946,
0.01825042724609375,
0.017705947875976562,
0.01735970878601074,
0.0172542667388916,
0.0173809871673584,
0.01710626792907715,
0.017366586685180666,
0.016324506759643555,
0.01620722579956055,
0.016196664810180665,
0.017458267211914062,
0.017969467163085937,
0.01775746726989746,
0.01832898712158203,
0.01716594886779785,
0.017232345581054687,
0.017241147994995116,
0.017067865371704103,
0.01635410499572754,
0.016157464981079103,
0.016191545486450194,
0.016296024322509765,
0.016228023529052733
]
},
"throughput": {
"unit": "tokens/s",
"value": 56.06200745950015
},
"energy": null,
"efficiency": null
}
}
}