{ "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_type": "gpt2", "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, "memory": true, "latency": 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": "0.3.1", "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": 899.715072, "max_global_vram": 1011.748864, "max_process_vram": 46960.484352, "max_reserved": 725.614592, "max_allocated": 666.339328 }, "latency": { "unit": "s", "count": 1, "total": 7.3818359375, "mean": 7.3818359375, "stdev": 0.0, "p50": 7.3818359375, "p90": 7.3818359375, "p95": 7.3818359375, "p99": 7.3818359375, "values": [ 7.3818359375 ] }, "throughput": null, "energy": null, "efficiency": null }, "prefill": { "memory": { "unit": "MB", "max_ram": 1053.958144, "max_global_vram": 1107.480576, "max_process_vram": 282581.46304, "max_reserved": 725.614592, "max_allocated": 513.035264 }, "latency": { "unit": "s", "count": 55, "total": 0.5072228775024413, "mean": 0.009222234136408027, "stdev": 0.0003977263765678904, "p50": 0.00903216552734375, "p90": 0.009661989593505858, "p95": 0.009986293983459472, "p99": 0.010618509216308594, "values": [ 0.00990320587158203, 0.010180166244506836, 0.009832165718078613, 0.009616485595703125, 0.00936096477508545, 0.00925808620452881, 0.009444326400756835, 0.00928752613067627, 0.009286405563354493, 0.00931280517578125, 0.009692325592041015, 0.009530566215515136, 0.009331046104431152, 0.009515365600585938, 0.009302725791931152, 0.009543524742126466, 0.010211206436157227, 0.011096647262573242, 0.008962725639343262, 0.008957765579223632, 0.009067206382751465, 0.00895024585723877, 0.00897376537322998, 0.009008966445922851, 0.008954886436462403, 0.009027046203613281, 0.009108965873718261, 0.00898912525177002, 0.009021445274353028, 0.0089814453125, 0.00905280590057373, 0.009018405914306641, 0.008985284805297852, 0.009014565467834473, 0.009034085273742675, 0.009008325576782227, 0.009060806274414062, 0.008992646217346191, 0.008981125831604004, 0.009001925468444824, 0.008992804527282714, 0.008990885734558106, 0.009140166282653808, 0.009114886283874512, 0.00902352523803711, 0.009130085945129394, 0.008964165687561036, 0.008954405784606934, 0.00903968620300293, 0.00903216552734375, 0.008999205589294434, 0.00899488639831543, 0.00901568603515625, 0.008980645179748534, 0.008990565299987792 ] }, "throughput": { "unit": "tokens/s", "value": 216.8671897088682 }, "energy": null, "efficiency": null }, "decode": { "memory": { "unit": "MB", "max_ram": 1053.958144, "max_global_vram": 1107.513344, "max_process_vram": 273321.275392, "max_reserved": 725.614592, "max_allocated": 513.259008 }, "latency": { "unit": "s", "count": 55, "total": 0.4940137395858764, "mean": 0.00898206799247048, "stdev": 0.00022890218677182523, "p50": 0.008938244819641113, "p90": 0.009215014076232911, "p95": 0.00955579776763916, "p99": 0.009743615856170654, "values": [ 0.009811526298522949, 0.009624006271362305, 0.009203206062316895, 0.009064005851745606, 0.008848325729370117, 0.009044005393981933, 0.008948165893554688, 0.00894384479522705, 0.008900164604187012, 0.009364645957946777, 0.009122566223144531, 0.00898320484161377, 0.008964486122131347, 0.008867685317993163, 0.008895365715026856, 0.009685766220092774, 0.009526565551757812, 0.008775364875793458, 0.00867360496520996, 0.008940164566040038, 0.008961605072021485, 0.009024966239929199, 0.009222886085510255, 0.008867204666137695, 0.008832485198974609, 0.008974245071411133, 0.008815365791320801, 0.00894592571258545, 0.008886885643005372, 0.00887008571624756, 0.009063685417175293, 0.008967524528503417, 0.0088436861038208, 0.008995205879211425, 0.008930886268615722, 0.009053924560546874, 0.008816164970397949, 0.008909284591674805, 0.008929286003112793, 0.008955685615539551, 0.008966245651245118, 0.008837605476379394, 0.009073445320129395, 0.00901296615600586, 0.008869926452636718, 0.008848484992980956, 0.008746404647827149, 0.008792004585266112, 0.008895845413208007, 0.008697284698486327, 0.008768964767456054, 0.0087620849609375, 0.008938244819641113, 0.008863045692443847, 0.008887525558471679 ] }, "throughput": { "unit": "tokens/s", "value": 111.33293589385913 }, "energy": null, "efficiency": null }, "per_token": { "memory": null, "latency": { "unit": "s", "count": 55, "total": 0.45135916233062745, "mean": 0.008206530224193227, "stdev": 0.00022735787258613434, "p50": 0.008119845390319824, "p90": 0.00848202133178711, "p95": 0.008800773048400878, "p99": 0.008928009071350098, "values": [ 0.009000325202941894, 0.00881600570678711, 0.008516965866088867, 0.00837568473815918, 0.008224005699157716, 0.008335365295410156, 0.008270565032958985, 0.008217124938964844, 0.008190884590148925, 0.008628005981445313, 0.008429604530334473, 0.008312965393066406, 0.008260644912719727, 0.00817792510986328, 0.008194725036621093, 0.008866406440734864, 0.008794244766235351, 0.008085764884948731, 0.008031044960021973, 0.008016324996948243, 0.008046725273132323, 0.008135205268859863, 0.008372164726257325, 0.008108004570007325, 0.007999684810638428, 0.00826640510559082, 0.008039845466613769, 0.008139044761657714, 0.008041125297546386, 0.008090246200561523, 0.008168805122375488, 0.008111044883728028, 0.00808464527130127, 0.008096324920654296, 0.008066884994506836, 0.008248004913330078, 0.008105605125427247, 0.008097604751586914, 0.008119845390319824, 0.008125124931335449, 0.008078885078430176, 0.008074085235595703, 0.008178085327148437, 0.008139525413513184, 0.008029125213623046, 0.008077605247497559, 0.008019044876098632, 0.00808800506591797, 0.008208325386047363, 0.007992644786834717, 0.008024165153503418, 0.008014565467834472, 0.008083683967590332, 0.008118885040283203, 0.008025605201721192 ] }, "throughput": { "unit": "tokens/s", "value": 121.85417864567832 }, "energy": null, "efficiency": null } } }