Fine-tuning
Benchmarks
DatasetSOTA methodMetricTrendResultsLast updated
81.1Accuracy
16
Jun 16, 2026
79.9Accuracy
16
Jun 16, 2026
94Accuracy
16
Jun 16, 2026
70.2Accuracy
16
Jun 16, 2026
2.5Test Loss
14
Mar 17, 2026
2.34Train Loss
14
Mar 17, 2026
69.12MMLU STEM Accuracy
14
Feb 26, 2026
1.93Test Loss
12
Jul 7, 2026
2.489Test Perplexity
12
May 7, 2026
0.335Training Loss
12
May 7, 2026
7.38Test Perplexity
12
May 7, 2026
0.562Loss
12
May 7, 2026
0.5Learning Score
8
Feb 26, 2026
0.661Evaluation Loss
7
Jun 1, 2026
0.773Evaluation Loss
7
Jun 1, 2026
0.706Evaluation Loss
7
Jun 1, 2026
0.638Evaluation Loss
7
Jun 1, 2026
1.35Evaluation Loss
7
Jun 1, 2026
7.42Weights Memory Usage
5
Mar 19, 2026
47,388Max Memory (MB)
4
Feb 26, 2026