diff --git a/README.md b/README.md index 831ad5f..1266b0e 100644 --- a/README.md +++ b/README.md @@ -31,11 +31,11 @@ Recently, lightweight Vision Transformers (ViTs) demonstrate superior performanc | Model | Top-1 (300 / 450)| #params | MACs | Latency | Ckpt | Core ML | Log | |:---------------|:----:|:---:|:--:|:--:|:--:|:--:|:--:| -| RepViT-M0.9 | 78.7 / 79.1 | 5.1M | 0.8G | 0.9ms | [M0.9-300e]() / [M0.9-450e]() | [M0.9-300e]() / [M0.9-450e]() | [M0.9-300e](./logs/repvit_m0_9_distill_300e.txt) / [M0.9-450e](./logs/repvit_m0_9_distill_450e.txt) | -| RepViT-M1.0 | 80.0 / 80.3 | 6.8M | 1.1G | 1.0ms | [M1.0-300e]() / [M1.0-450e]() | [M1.0-300e]() / [M1.0-450e]() | [M1.0-300e](./logs/repvit_m1_0_distill_300e.txt) / [M1.0-450e](./logs/repvit_m1_0_distill_450e.txt) | -| RepViT-M1.1 | 80.7 / 81.1 | 8.2M | 1.3G | 1.1ms | [M1.1-300e]() / [M1.1-450e]() | [M1.1-300e]() / [M1.1-450e]() | [M1.1-300e](./logs/repvit_m1_1_distill_300e.txt) / [M1.1-450e](./logs/repvit_m1_1_distill_450e.txt) | -| RepViT-M1.5 | 82.3 / 82.5 | 14.0M | 2.3G | 1.5ms | [M1.5-300e]() / [M1.5-450e]() | [M1.5-300e]() / [M1.5-450e]() | [M1.5-300e](./logs/repvit_m1_5_distill_300e.txt) / [M1.5-450e](./logs/repvit_m1_5_distill_450e.txt) | -| RepViT-M2.3 | 83.3 / 83.7 | 22.9M | 4.5G | 2.3ms | [M2.3-300e]() / [M2.3-450e]() | [M2.3-300e]() / [M2.3-450e]() | [M2.3-300e](./logs/repvit_m2_3_distill_300e.txt) / [M2.3-450e](./logs/repvit_m2_3_distill_450e.txt) | +| M0.9 | 78.7 / 79.1 | 5.1M | 0.8G | 0.9ms | [300e]() / [450e]() | [300e]() / [450e]() | [300e](./logs/repvit_m0_9_distill_300e.txt) / [450e](./logs/repvit_m0_9_distill_450e.txt) | +| M1.0 | 80.0 / 80.3 | 6.8M | 1.1G | 1.0ms | [300e]() / [450e]() | [300e]() / [450e]() | [300e](./logs/repvit_m1_0_distill_300e.txt) / [450e](./logs/repvit_m1_0_distill_450e.txt) | +| M1.1 | 80.7 / 81.1 | 8.2M | 1.3G | 1.1ms | [300e]() / [450e]() | [300e]() / [450e]() | [300e](./logs/repvit_m1_1_distill_300e.txt) / [450e](./logs/repvit_m1_1_distill_450e.txt) | +| M1.5 | 82.3 / 82.5 | 14.0M | 2.3G | 1.5ms | [300e]() / [450e]() | [300e]() / [450e]() | [300e](./logs/repvit_m1_5_distill_300e.txt) / [450e](./logs/repvit_m1_5_distill_450e.txt) | +| M2.3 | 83.3 / 83.7 | 22.9M | 4.5G | 2.3ms | [300e]() / [450e]() | [300e]() / [450e]() | [300e](./logs/repvit_m2_3_distill_300e.txt) / [450e](./logs/repvit_m2_3_distill_450e.txt) | Tips: Convert a training-time RepViT into the inference-time structure diff --git a/detection/README.md b/detection/README.md index df829a5..cb7a420 100644 --- a/detection/README.md +++ b/detection/README.md @@ -6,8 +6,8 @@ Detection and instance segmentation on MS COCO 2017 is implemented based on [MMD | Model | $AP^b$ | $AP_{50}^b$ | $AP_{75}^b$ | $AP^m$ | $AP_{50}^m$ | $AP_{75}^m$ | Latency | Ckpt | Log | |:---------------|:----:|:---:|:--:|:--:|:--:|:--:|:--:|:--:|:--:| | RepViT-M1_1 | 39.8 | 61.9 | 43.5 | 37.2 | 58.8 | 40.1 | 4.9ms | [M1_1]() | [M1_1](./logs/repvit_m1_1_coco.json) | -| RepViT-M1_5 | 41.6 | 63.2 | 45.3 | 38.6 | 60.5 | 41.5 | 43.6 | 6.4ms | [M1_5]() | [M1_5](./logs/repvit_m1_5_coco.json) | -| RepViT-M2_3 | 44.6 | 66.1 | 48.8 | 40.8 | 63.6 | 43.9 | 46.1 | 9.9ms | [M2_3]() | [M2_3](./logs/repvit_m2_3_coco.json) | +| RepViT-M1_5 | 41.6 | 63.2 | 45.3 | 38.6 | 60.5 | 41.5 | 6.4ms | [M1_5]() | [M1_5](./logs/repvit_m1_5_coco.json) | +| RepViT-M2_3 | 44.6 | 66.1 | 48.8 | 40.8 | 63.6 | 43.9 | 9.9ms | [M2_3]() | [M2_3](./logs/repvit_m2_3_coco.json) | ## Installation diff --git a/figures/latency.png b/figures/latency.png index 059a05b..fb804fd 100644 Binary files a/figures/latency.png and b/figures/latency.png differ diff --git a/figures/repvit_m1_latency.png b/figures/repvit_m0_9_latency.png similarity index 100% rename from figures/repvit_m1_latency.png rename to figures/repvit_m0_9_latency.png