LMM Inference Efficiency on 8x RTX 4090 GPUs
1.93FLOPs (T)
LLaVA-FA-2B
Updated 4d ago
Evaluation Results
| Method | Links | ||
|---|---|---|---|
2026.01 | 1.93 | 30.6 | |
2026.01 | 2.01 | 35.1 | |
2026.01 | 2.14 | 36.2 | |
2026.01 | 2.48 | 39.3 | |
2026.01 | 2.81 | 40.1 | |
2026.01 | 2.95 | 41.5 |
LLaVA-FA-2B
| Method | Links | ||
|---|---|---|---|
2026.01 | 1.93 | 30.6 | |
2026.01 | 2.01 | 35.1 | |
2026.01 | 2.14 | 36.2 | |
2026.01 | 2.48 | 39.3 | |
2026.01 | 2.81 | 40.1 | |
2026.01 | 2.95 | 41.5 |