Benchmarks / results_virtuoso_large.py
Julien Simon
Add g6e.48xlarge
f011c00
raw
history blame contribute delete
441 Bytes
"""Module containing performance results for the Virtuoso-Large model."""
results_virtuoso_large = {
"name": "Virtuoso-Large",
"modelType": "Qwen2 72B", # 2.5
"configurations": [
{
"instanceType": "g6e.48xlarge",
"quantization": "none",
"container": "LMI 0.30+vLLM 0.6.2",
"status": "OK",
"tokensPerSecond": "28",
"notes": "",
},
],
}