Jae-Won Chung
New leaderboard prototype
b10121d
raw
history blame contribute delete
322 Bytes
{
"Model": "codellama/CodeLlama-7b-hf",
"GPU": "NVIDIA A100-SXM4-40GB",
"TP": 1,
"PP": 1,
"Energy/req (J)": 28.59409628402536,
"Avg TPOT (s)": 0.037676650237406725,
"Token tput (tok/s)": 834.3207422168283,
"Avg Output Tokens": 99.17530487804878,
"Avg BS (reqs)": 31.7431633698479,
"Max BS (reqs)": 32
}