Language Modeling on Nemotron-3-Nano (val)
1.4122Validation LossMuon-Jordan
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Muon-JordanModel=4B-A770M2026.05 | 1.4122 | 4.1052 | |
| cubic5Model=4B-A770M2026.05 | 1.4132 | 4.1093 | |
| cubic5Model=2B-A500M2026.05 | 1.4967 | 4.4669 | |
| Muon-JordanModel=2B-A500M2026.05 | 1.4968 | 4.4673 | |
| cubic5Model=1B-A315M2026.05 | 1.6087 | 4.9965 | |
| Muon-JordanModel=1B-A315M2026.05 | 1.6095 | 5.0004 |