|
{ |
|
"base_current_gpu_type": "NVIDIA A100-PCIE-40GB", |
|
"base_current_gpu_total_memory": 40339.3125, |
|
"base_token_generation_latency_sync": 32.622706985473634, |
|
"base_token_generation_latency_async": 33.04024673998356, |
|
"base_token_generation_throughput_sync": 0.03065349544552768, |
|
"base_token_generation_throughput_async": 0.030266117800804827, |
|
"base_token_generation_CO2_emissions": null, |
|
"base_token_generation_energy_consumption": null, |
|
"base_inference_latency_sync": 32.64716815948486, |
|
"base_inference_latency_async": 32.19940662384033, |
|
"base_inference_throughput_sync": 0.03063052804809576, |
|
"base_inference_throughput_async": 0.03105647292455394, |
|
"base_inference_CO2_emissions": null, |
|
"base_inference_energy_consumption": null, |
|
"smashed_current_gpu_type": "NVIDIA A100-PCIE-40GB", |
|
"smashed_current_gpu_total_memory": 40339.3125, |
|
"smashed_token_generation_latency_sync": 65.22703285217285, |
|
"smashed_token_generation_latency_async": 64.26504217088223, |
|
"smashed_token_generation_throughput_sync": 0.015331066833384064, |
|
"smashed_token_generation_throughput_async": 0.015560559305959483, |
|
"smashed_token_generation_CO2_emissions": null, |
|
"smashed_token_generation_energy_consumption": null, |
|
"smashed_inference_latency_sync": 64.54333419799805, |
|
"smashed_inference_latency_async": 62.95466423034668, |
|
"smashed_inference_throughput_sync": 0.015493466713887507, |
|
"smashed_inference_throughput_async": 0.015884446565246867, |
|
"smashed_inference_CO2_emissions": null, |
|
"smashed_inference_energy_consumption": null |
|
} |