Language Modeling on Set (val)
1.61LossEngram-40B
Evaluation results
| Method | Links | |
|---|---|---|
| Engram-40BShots=-2026.01 | 1.61 | |
| Engram-27BShots=-2026.01 | 1.622 | |
| MoE-27BShots=-2026.01 | 1.634 | |
| Dense-4BShots=-2026.01 | 1.768 |
| Method | Links | |
|---|---|---|
| Engram-40BShots=-2026.01 | 1.61 | |
| Engram-27BShots=-2026.01 | 1.622 | |
| MoE-27BShots=-2026.01 | 1.634 | |
| Dense-4BShots=-2026.01 | 1.768 |