Table 6 Computational efficiency summary.
Metric | Value |
|---|---|
Total parameters | 62,995,651 (~ 240.31 MB) |
Trainable parameters | 42,643,580 (~ 162.67 MB) |
Non-trainable parameters | 20,352,071 (~ 77.64 MB) |
Number of layers | 3 base models + weighted summation |
Input shape | (224, 224, 3) |
Output shape | (None, 4) |
Memory usage | ~ 240.31 MB |
Total training time | 7,637Â s |
Average step time | 0.3577Â s |
Validation time | 69Â s |
Test time | 31Â s |
