Here, we give the full list of publicly pre-trained models supported by the Hailo Model Zoo.
- Benchmark Networks are marked with
- Networks available in TAPPAS are marked with
- Benchmark and TAPPAS networks run in performance mode
- All models were compiled using Hailo Dataflow Compiler v3.30.0
| Network Name | Accuracy (top1) | HW Accuracy | FPS (Batch Size=1) | FPS (Batch Size=8) | Input Resolution (HxWxC) | Params (M) | OPS (G) | Pretrained | Source | Compiled | Profile Html |
|---|---|---|---|---|---|---|---|---|---|---|---|
| cas_vit_m | 81.2 | 80.97 | 38 | 71 | 384x384x3 | 12.42 | 10.89 | download | link | download | download |
| cas_vit_s | 79.93 | 79.83 | 52 | 103 | 384x384x3 | 5.5 | 5.4 | download | link | download | download |
| cas_vit_t | 81.9 | 81.63 | 26 | 39 | 384x384x3 | 21.76 | 20.85 | download | link | download | download |
| davit_tiny | 82.7 | 82.18 | 0 | 0 | 224x224x3 | 28.36 | 9.1 | download | link | download | download |
| deit_base | 80.93 | 79.78 | 29 | 60 | 224x224x3 | 80.26 | 35.22 | download | link | download | download |
| deit_small | 78.25 | 77.52 | 46 | 97 | 224x224x3 | 20.52 | 9.4 | download | link | download | download |
| deit_tiny | 69.07 | 68.69 | 83 | 237 | 224x224x3 | 5.3 | 2.57 | download | link | download | download |
| efficientformer_l1 | 79.13 | 76.57 | 45 | 66 | 224x224x3 | 12.3 | 2.6 | download | link | download | download |
| efficientnet_l | 80.47 | 79.29 | 85 | 163 | 300x300x3 | 10.55 | 19.4 | download | link | download | download |
| efficientnet_lite0 | 74.99 | 73.82 | 301 | 765 | 224x224x3 | 4.63 | 0.78 | download | link | download | download |
| efficientnet_lite1 | 76.67 | 76.27 | 215 | 513 | 240x240x3 | 5.39 | 1.22 | download | link | download | download |
| efficientnet_lite2 | 77.46 | 76.76 | 149 | 322 | 260x260x3 | 6.06 | 1.74 | download | link | download | download |
| efficientnet_lite3 | 79.29 | 78.79 | 109 | 231 | 280x280x3 | 8.16 | 2.8 | download | link | download | download |
| efficientnet_lite4 | 80.79 | 80.06 | 77 | 153 | 300x300x3 | 12.95 | 5.10 | download | link | download | download |
| efficientnet_m | 78.91 | 78.53 | 140 | 268 | 240x240x3 | 6.87 | 7.32 | download | link | download | download |
| efficientnet_s | 77.63 | 77.27 | 189 | 383 | 224x224x3 | 5.41 | 4.72 | download | link | download | download |
| fastvit_sa12 | 79.8 | 76.85 | 165 | 436 | 224x224x3 | 11.99 | 3.59 | download | link | download | download |
| hardnet39ds | 73.43 | 73.01 | 319 | 836 | 224x224x3 | 3.48 | 0.86 | download | link | download | download |
| hardnet68 | 75.47 | 75.25 | 134 | 263 | 224x224x3 | 17.56 | 8.5 | download | link | download | download |
| inception_v1 | 69.74 | 69.54 | 272 | 556 | 224x224x3 | 6.62 | 3 | download | link | download | download |
| mobilenet_v1 | 70.97 | 70.3 | 1426 | 1426 | 224x224x3 | 4.22 | 1.14 | download | link | download | download |
mobilenet_v2_1.0  |
71.78 | 70.85 | 869 | 869 | 224x224x3 | 3.49 | 0.62 | download | link | download | download |
| mobilenet_v2_1.4 | 74.18 | 73.2 | 294 | 673 | 224x224x3 | 6.09 | 1.18 | download | link | download | download |
| mobilenet_v3 | 72.21 | 71.81 | 346 | 793 | 224x224x3 | 4.07 | 2 | download | link | download | download |
| mobilenet_v3_large_minimalistic | 72.12 | 70.6 | 492 | 1309 | 224x224x3 | 3.91 | 0.42 | download | link | download | download |
| regnetx_1.6gf | 77.05 | 76.77 | 323 | 790 | 224x224x3 | 9.17 | 3.22 | download | link | download | download |
| regnetx_800mf | 75.16 | 74.87 | 459 | 1195 | 224x224x3 | 7.24 | 1.6 | download | link | download | download |
| repghost_1_0x | 73.03 | 72.2 | 208 | 489 | 224x224x3 | 4.1 | 0.28 | download | link | download | download |
| repghost_2_0x | 77.18 | 76.93 | 130 | 316 | 224x224x3 | 9.8 | 1.04 | download | link | download | download |
| repvgg_a1 | 74.4 | 72.53 | 312 | 643 | 224x224x3 | 12.79 | 4.7 | download | link | download | download |
| repvgg_a2 | 76.52 | 74.47 | 190 | 336 | 224x224x3 | 25.5 | 10.2 | download | link | download | download |
| resmlp12_relu | 75.27 | 74.82 | 89 | 311 | 224x224x3 | 15.77 | 6.04 | download | link | download | download |
| resnet_v1_18 | 71.27 | 70.79 | 416 | 865 | 224x224x3 | 11.68 | 3.64 | download | link | download | download |
| resnet_v1_34 | 72.7 | 72.18 | 218 | 477 | 224x224x3 | 21.79 | 7.34 | download | link | download | download |
resnet_v1_50  |
75.21 | 74.67 | 222 | 510 | 224x224x3 | 25.53 | 6.98 | download | link | download | download |
| resnext26_32x4d | 76.17 | 75.94 | 263 | 508 | 224x224x3 | 15.37 | 4.96 | download | link | download | download |
| resnext50_32x4d | 79.3 | 78.4 | 158 | 333 | 224x224x3 | 24.99 | 8.48 | download | link | download | download |
| squeezenet_v1.1 | 59.85 | 59.35 | 711 | 1287 | 224x224x3 | 1.24 | 0.78 | download | link | download | download |
| swin_small | 83.13 | 80.25 | 13 | 27 | 224x224x3 | 50 | 17.6 | download | link | download | download |
| swin_tiny | 81.3 | 79.54 | 26 | 48 | 224x224x3 | 29 | 9.1 | download | link | download | download |
| vit_base | 84.5 | 83.45 | 29 | 60 | 224x224x3 | 86.5 | 35.188 | download | link | download | download |
| vit_base_bn |
79.98 | 79.24 | 52 | 135 | 224x224x3 | 86.5 | 35.188 | download | link | download | download |
| vit_small | 81.5 | 80.38 | 52 | 119 | 224x224x3 | 21.12 | 8.62 | download | link | download | download |
| vit_small_bn | 78.12 | 77.26 | 116 | 337 | 224x224x3 | 21.12 | 8.62 | download | link | download | download |
| vit_tiny | 75.51 | 74.15 | 87 | 263 | 224x224x3 | 5.73 | 2.2 | download | link | download | download |
| vit_tiny_bn | 68.95 | 67.33 | 206 | 800 | 224x224x3 | 5.73 | 2.2 | download | link | download | download |