|
{ |
|
"base_current_gpu_type": "NVIDIA A100-PCIE-40GB", |
|
"base_current_gpu_total_memory": 40339.3125, |
|
"base_token_generation_latency_sync": 33.9055233001709, |
|
"base_token_generation_latency_async": 33.37778877466917, |
|
"base_token_generation_throughput_sync": 0.02949371968533987, |
|
"base_token_generation_throughput_async": 0.029960043391457757, |
|
"base_token_generation_CO2_emissions": null, |
|
"base_token_generation_energy_consumption": null, |
|
"base_inference_latency_sync": 32.31610870361328, |
|
"base_inference_latency_async": 30.997133255004883, |
|
"base_inference_throughput_sync": 0.030944319725233177, |
|
"base_inference_throughput_async": 0.032261047877339986, |
|
"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": 74.24477767944336, |
|
"smashed_token_generation_latency_async": 73.8770842552185, |
|
"smashed_token_generation_throughput_sync": 0.013468960797721892, |
|
"smashed_token_generation_throughput_async": 0.013535997123889771, |
|
"smashed_token_generation_CO2_emissions": null, |
|
"smashed_token_generation_energy_consumption": null, |
|
"smashed_inference_latency_sync": 70.68764038085938, |
|
"smashed_inference_latency_async": 69.0847396850586, |
|
"smashed_inference_throughput_sync": 0.014146744672931217, |
|
"smashed_inference_throughput_async": 0.014474976739563173, |
|
"smashed_inference_CO2_emissions": null, |
|
"smashed_inference_energy_consumption": null |
|
} |