Text Generation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
71.93Accuracy
5
Jul 7, 2026
67.5Median Adv
5
Jul 7, 2026
25.876Perplexity
5
Jul 2, 2026
32.76Decode tok/s
5
Jun 4, 2026
32.94BLEU-4
5
May 26, 2026
0.78Spearman Correlation (ρ)
5
May 19, 2026
42Forget Gen.
5
May 18, 2026
100ROUGE-1
5
May 12, 2026
6.48NLL
5
Mar 23, 2026
0.8Throughput (K tok/s)
5
Feb 26, 2026
28.69BLEU
5
Feb 26, 2026
1.23SacreBLEU
5
Feb 26, 2026
22.47ROUGE-L
5
Feb 26, 2026
20.93PPL
5
Feb 26, 2026
12.16BLEU
4
May 28, 2026
80TTR
4
May 22, 2026
59.9MAUVE
4
May 18, 2026
0.555BERTScore
4
May 12, 2026
38.64TPOT (ms/token)
4
May 8, 2026
25.52TPOT (ms/token)
4
May 8, 2026
24.03Latency (ms/token)
4
May 8, 2026
23.97TPOT (ms/token)
4
May 8, 2026
24.14TPOT (ms/token)
4
May 8, 2026
24.15TPOT (ms/token)
4
May 8, 2026
24.11TPOT (ms/token)
4
May 8, 2026