speeds updated from FP32 to FP16

README.md

@@ -4,7 +4,7 @@
 
 This repository represents Ultralytics open-source research into future object detection methods, and incorporates our lessons learned and best practices evolved over training thousands of models on custom client datasets with our previous YOLO repository https://github.com/ultralytics/yolov3. **All code and models are under active development, and are subject to modification or deletion without notice.** Use at your own risk.
 
-<img src="https://user-images.githubusercontent.com/26833433/84200349-729f2680-aa5b-11ea-8f9a-604c9e01a658.png" width="1000">** GPU Latency measures end-to-end latency per image averaged over 5000 COCO val2017 images using a V100 GPU with batch size 32, and includes image preprocessing, PyTorch FP32 inference, postprocessing and NMS.
+<img src="https://user-images.githubusercontent.com/26833433/84200349-729f2680-aa5b-11ea-8f9a-604c9e01a658.png" width="1000">** GPU Speed measures end-to-end time per image averaged over 5000 COCO val2017 images using a V100 GPU with batch size 32, and includes image preprocessing, PyTorch FP32 inference, postprocessing and NMS.
 
 - **June 9, 2020**: [CSP](https://github.com/WongKinYiu/CrossStagePartialNetworks) updates to all YOLOv5 models. New models are faster, smaller and more accurate. Credit to @WongKinYiu for his excellent work with CSP. 
 - **May 27, 2020**: Public release of repo. YOLOv5 models are SOTA among all known YOLO implementations, YOLOv5 family will be undergoing architecture research and development over Q2/Q3 2020 to increase performance. Updates may include [CSP](https://github.com/WongKinYiu/CrossStagePartialNetworks) bottlenecks, [YOLOv4](https://github.com/AlexeyAB/darknet) features, as well as PANet or BiFPN heads.
@@ -13,17 +13,17 @@ This repository represents Ultralytics open-source research into future object d
 
 ## Pretrained Checkpoints
 
-| Model | AP<sup>val</sup> | AP<sup>test</sup> | AP<sub>50</sub> | Latency<sub>GPU</sub> | FPS<sub>GPU</sub> || params | FLOPs |
+| Model | AP<sup>val</sup> | AP<sup>test</sup> | AP<sub>50</sub> | Speed<sub>GPU</sub> | FPS<sub>GPU</sub> || params | FLOPs |
 |---------- |------ |------ |------ | -------- | ------| ------ |------  |  :------: |
-| YOLOv5-s ([ckpt](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J))    | 35.5     | 35.5     | 55.0     | **2.5ms** | **400** || 7.1M   | 12.6B
-| YOLOv5-m ([ckpt](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J))    | 42.7     | 42.7     | 62.4     | 4.4ms     | 227     || 22.0M  | 39.0B
-| YOLOv5-l ([ckpt](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J))    | 45.7     | 45.9     | 65.1     | 6.8ms     | 147     || 50.3M  | 89.0B
-| YOLOv5-x ([ckpt](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J))    | **47.2** | **47.3** | **66.6** | 11.7ms    | 85      || 95.9M | 170.3B
-| YOLOv3-SPP ([ckpt](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J))  | 45.6     | 45.5     | 65.2     | 7.9ms     | 127     || 63.0M  | 118.0B
+| YOLOv5-s ([ckpt](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J))    | 35.5     | 35.5     | 55.0     | **2.1ms** | **476** || 7.1M   | 12.6B
+| YOLOv5-m ([ckpt](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J))    | 42.7     | 42.7     | 62.4     | 3.2ms     | 312     || 22.0M  | 39.0B
+| YOLOv5-l ([ckpt](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J))    | 45.7     | 45.9     | 65.1     | 4.1ms     | 243     || 50.3M  | 89.0B
+| YOLOv5-x ([ckpt](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J))    | **47.2** | **47.3** | **66.6** | 6.5ms     | 153     || 95.9M  | 170.3B
+| YOLOv3-SPP ([ckpt](https://drive.google.com/open?id=1Drs_Aiu7xx6S-ix95f9kNsA6ueKRpN2J))  | 45.6     | 45.5     | 65.2     | 4.8ms     | 208     || 63.0M  | 118.0B
 
 ** AP<sup>test</sup> denotes COCO [test-dev2017](http://cocodataset.org/#upload) server results, all other AP results in the table denote val2017 accuracy.  
 ** All AP numbers are for single-model single-scale without ensemble or test-time augmentation. Reproduce by `python test.py --img 736 --conf 0.001`  
-** Latency<sub>GPU</sub> measures end-to-end latency per image averaged over 5000 COCO val2017 images using a GCP [n1-standard-16](https://cloud.google.com/compute/docs/machine-types#n1_standard_machine_types) instance with one V100 GPU, and includes image preprocessing, PyTorch FP32 inference at batch size 32, postprocessing and NMS. Average NMS time included in this chart is 1-2ms/img.  Reproduce by `python test.py --img 640 --conf 0.1`  
+** Speed<sub>GPU</sub> measures end-to-end time per image averaged over 5000 COCO val2017 images using a GCP [n1-standard-16](https://cloud.google.com/compute/docs/machine-types#n1_standard_machine_types) instance with one V100 GPU, and includes image preprocessing, PyTorch FP16 image inference at --batch-size 32 --img-size 640, postprocessing and NMS. Average NMS time included in this chart is 1-2ms/img.  Reproduce by `python test.py --img 640 --conf 0.1`  
 ** All checkpoints are trained to 300 epochs with default settings and hyperparameters (no autoaugmentation).