IlyasMoutawwakil's picture
Upload cpu_inference_transformers_text-generation_openai-community/gpt2/benchmark.json with huggingface_hub
11545f7 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": 16757.342208,
"system": "Linux",
"machine": "x86_64",
"platform": "Linux-6.5.0-1018-azure-x86_64-with-glibc2.35",
"processor": "x86_64",
"python_version": "3.10.14",
"optimum_benchmark_version": "0.2.0",
"optimum_benchmark_commit": "a8ccb91469272f9f38883fbf3e14bffcd5e95f2d",
"transformers_version": "4.40.2",
"transformers_commit": null,
"accelerate_version": "0.30.0",
"accelerate_commit": null,
"diffusers_version": "0.27.2",
"diffusers_commit": null,
"optimum_version": null,
"optimum_commit": null,
"timm_version": "0.9.16",
"timm_commit": null,
"peft_version": null,
"peft_commit": null
}
},
"report": {
"prefill": {
"memory": {
"unit": "MB",
"max_ram": 959.721472,
"max_global_vram": null,
"max_process_vram": null,
"max_reserved": null,
"max_allocated": null
},
"latency": {
"unit": "s",
"count": 13,
"total": 0.6312671800000658,
"mean": 0.048559013846158905,
"stdev": 0.0023036918286531358,
"p50": 0.049563321999983145,
"p90": 0.05037565600000562,
"p95": 0.050465417600003094,
"p99": 0.05057309071999498,
"values": [
0.04530646500001012,
0.04400738800001136,
0.044212981999976364,
0.04959109300000364,
0.05037551999998868,
0.0486010739999756,
0.050375690000009854,
0.05033551500002886,
0.05002007500002037,
0.049563321999983145,
0.05060000899999295,
0.0491694050000433,
0.04910864200002152
]
},
"throughput": {
"unit": "tokens/s",
"value": 41.186997873067455
},
"energy": {
"unit": "kWh",
"cpu": 1.6731792577990781e-06,
"ram": 6.992546118709696e-08,
"gpu": 0.0,
"total": 1.743104718986175e-06
},
"efficiency": {
"unit": "tokens/kWh",
"value": 1147377.9964081794
}
},
"decode": {
"memory": {
"unit": "MB",
"max_ram": 959.721472,
"max_global_vram": null,
"max_process_vram": null,
"max_reserved": null,
"max_allocated": null
},
"latency": {
"unit": "s",
"count": 13,
"total": 0.39328847399991673,
"mean": 0.030252959538455134,
"stdev": 0.0013191504688200174,
"p50": 0.030537863999995807,
"p90": 0.03161084880000544,
"p95": 0.031753168599993845,
"p99": 0.03188260731998071,
"values": [
0.027211818000012045,
0.028034694999973908,
0.030499201999987235,
0.030537863999995807,
0.029951227000026392,
0.03191496699997742,
0.031080486999996992,
0.02936419999997497,
0.03006252500000528,
0.03164530300000479,
0.03067822499997419,
0.03083492899997964,
0.03147303200000806
]
},
"throughput": {
"unit": "tokens/s",
"value": 33.05461730872579
},
"energy": {
"unit": "kWh",
"cpu": 9.73473069844423e-07,
"ram": 4.0683460135085154e-08,
"gpu": 0.0,
"total": 1.0141565299795071e-06
},
"efficiency": {
"unit": "tokens/kWh",
"value": 986041.0798914905
}
},
"per_token": {
"memory": null,
"latency": {
"unit": "s",
"count": 12,
"total": 0.9520682520000037,
"mean": 0.0793390210000003,
"stdev": 0.0027901983800490988,
"p50": 0.08025918549998323,
"p90": 0.08127993490002723,
"p95": 0.08139592595002795,
"p99": 0.08150457559002461,
"values": [
0.07208035500002552,
0.07455751499998087,
0.08014541799997232,
0.08037295299999414,
0.0804896320000239,
0.08153173800002378,
0.0797304429999599,
0.08004407900000388,
0.08123608599998988,
0.08128480700003138,
0.08002632599999515,
0.08056890000000294
]
},
"throughput": {
"unit": "tokens/s",
"value": 12.604138384818187
},
"energy": null,
"efficiency": null
}
}
}