Training Efficiency on Mixtral-8x22B Coarse-grained
49.3MFU
MCore w/ Folding
Updated 1mo ago
Evaluation Results
| Method | Links | |
|---|---|---|
2025.04 | 49.3 | |
2025.04 | 46.3 | |
2025.04 | 36.6 | |
2025.04 | 23.4 | |
2025.04 | 4.3 |
MCore w/ Folding
| Method | Links | |
|---|---|---|
2025.04 | 49.3 | |
2025.04 | 46.3 | |
2025.04 | 36.6 | |
2025.04 | 23.4 | |
2025.04 | 4.3 |