Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

GPT-J

Benchmarks

Task NameDataset NameSOTA ResultTrend
Inference ThroughputGPT-J 6B
Inference Throughput (tokens/s)1,484
16
Inference LatencyGPT-J-6B
Original Inference Time (s)0.203
8
Jailbreak AttackGPT-J unseen instances
Hypervolume0.61
7
Showing 3 of 3 rows