Language Modeling
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
2.942CC Loss
4
Jun 8, 2026
28.52Perplexity (1.3B tokens)
4
Jun 2, 2026
2.434Validation Loss
4
Jun 2, 2026
8.59Perplexity (Dense)
4
May 28, 2026
3.261Loss
4
May 22, 2026
3.238Loss
4
May 22, 2026
0.678Avg Delta PPL
4
May 21, 2026
17.42Perplexity (PPL)
4
May 19, 2026
0.977BPB
4
May 15, 2026
21.9Validation Perplexity
4
May 13, 2026
12.24Perplexity
4
May 12, 2026
37.05Perplexity
4
May 12, 2026
2.3Validation Loss
4
Jul 3, 2026
2.1Train Loss
4
Jul 3, 2026
191.3Tokens Per Second (TPS)
4
May 7, 2026
216.6Throughput (tok/s)
4
May 7, 2026
3.079Validation Loss
4
May 6, 2026
3.37Validation Loss
4
May 6, 2026
0.151Loss
4
Apr 28, 2026
237.1Perplexity (PPL)
4
Apr 28, 2026
8.66Perplexity
4
Apr 21, 2026
5.48Perplexity
4
Apr 21, 2026
1.122BPC
4
Jun 18, 2026
47.61Perplexity (PPL)
4
Apr 21, 2026
10.09Perplexity
4
Apr 15, 2026