From d4b80b82c30dfb7b249fe8f80a5a9f11acf1ba14 Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 16 Apr 2019 14:01:55 +0200 Subject: [PATCH 1/2] Update README.md --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 7a149011..b6afbd99 100755 --- a/README.md +++ b/README.md @@ -52,7 +52,7 @@ Each epoch trains on 117,263 images from the train and validate COCO sets, and t Here we see training results from `coco_1img.data`, `coco_10img.data` and `coco_100img.data`, 3 example files available in the `data/` folder, which train and test on the first 1, 10 and 100 images of the coco2014 trainval dataset. `from utils import utils; utils.plot_results()` -![results](https://user-images.githubusercontent.com/26833433/55669383-df76c980-5876-11e9-9806-691bd507ee17.jpg) +![results](https://user-images.githubusercontent.com/26833433/56207787-ec9e7000-604f-11e9-94dd-e1fcc374270f.png) ## Image Augmentation From ddc3c82c91a92a4e3121c416653d63daea80c6af Mon Sep 17 00:00:00 2001 From: Glenn Jocher Date: Tue, 16 Apr 2019 22:29:00 +0200 Subject: [PATCH 2/2] Update README.md --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index b6afbd99..dd1b2fcf 100755 --- a/README.md +++ b/README.md @@ -47,7 +47,7 @@ Python 3.7 or later with the following `pip3 install -U -r requirements.txt` pac **Resume Training:** Run `train.py --resume` resumes training from the latest checkpoint `weights/latest.pt`. -Each epoch trains on 117,263 images from the train and validate COCO sets, and tests on 5000 images from the COCO validate set. Default training settings produce loss plots below, with **training speed of 0.6 s/batch on a 1080 Ti (18 epochs/day)** or 0.45 s/batch on a 2080 Ti. +Each epoch trains on 117,263 images from the train and validate COCO sets, and tests on 5000 images from the COCO validate set. Default training settings produce loss plots below, with **training speed of 0.25 s/batch on a V100 GPU (almost 50 COCO epochs/day)**. Here we see training results from `coco_1img.data`, `coco_10img.data` and `coco_100img.data`, 3 example files available in the `data/` folder, which train and test on the first 1, 10 and 100 images of the coco2014 trainval dataset.