much needed update of readmes

wjtscn · Nov 4, 2019 · 4acd3ae · 4acd3ae
1 parent 3d65d26
commit 4acd3ae
Show file tree

Hide file tree

Showing 2 changed files with 11 additions and 74 deletions.
diff --git a/README.md b/README.md
@@ -87,40 +87,37 @@ conda activate viddet-mx
 ### Training
 To train a model you can use something like:
 ```
-python train_yolov3.py --dataset voc --gpus 0,1,2,3 --save_prefix 0001 --num_workers 16 --warmup_lr 0.0001 --warmup_epochs 3 --syncbn
+python train_yolov3.py --dataset voc --gpus 0,1,2,3 --save_prefix 0001 --warmup_epochs 3 --syncbn
 ```
 
-If you don't have this much power available you will need to specify a lower batch size:
+If you don't have this much power available you will need to specify a lower batch size (this also will default to one GPU):
 ```
-python train_yolov3.py --batch_size 4 --dataset voc --save_prefix 0001 --warmup_lr 0.0001 --warmup_epochs 3
+python train_yolov3.py --batch_size 4 --dataset voc --save_prefix 0001 --warmup_epochs 3
 ```
 
-Using MobileNet1.0 `--network mobilenet1.0` and batch size of 16
-`--batch_size 16` for `voc` uses **9GB** of GPU memory
-
 ***We found a warmup was necessary for YOLOv3***
 
 ### Finetuning
 To finetune a model you need to specify a `--resume` path to a
  pretrained params model file and specify the `--trained_on` dataset,
  the model will be finetuned on the dataset specified with `--dataset`
 ```
-python train_yolov3.py --dataset voc --trained_on coco --resume models/0003/yolo3_darknet53_coco_best.params --gpus 0,1,2,3 --save_prefix 0006 --num_workers 16 --warmup_lr 0.0001 --warmup_epochs 3 --syncbn
+python train_yolov3.py --dataset voc --trained_on coco --resume models/experiments/0003/yolo3_darknet53_coco_best.params --gpus 0,1,2,3 --save_prefix 0006 --warmup_epochs 3 --syncbn
 ```
 
 ### Detection, Testing & Visualisation
 To evaluate a model you can use something like:
 ```
-python detect_yolov3.py --batch_size 1 --model_path models/0001/yolo3_darknet53_voc_best.params --metrics voc --dataset voc --save_prefix 0001
+python detect_yolov3.py --batch_size 1 --model_path models/experiments/0001/yolo3_darknet53_voc_best.params --metrics voc --dataset voc --save_prefix 0001
 ```
 
 You can also evaluate on different data than the model was trained on
 (voc trained model on vid set):
 ```
-python detect_yolov3.py --trained_on voc --batch_size 1 --model_path models/0001/yolo3_darknet53_voc_best.params --metrics voc,coco,vid --dataset vid --save_prefix 0001
+python detect_yolov3.py --trained_on voc --batch_size 1 --model_path models/experiments/0001/yolo3_darknet53_voc_best.params --metrics voc,coco,vid --dataset vid --save_prefix 0001
 ```
 
-Visualisation is **off** by default use `--visualise True` to write out images with boxes displayed.
+Visualisation is **off** by default add `--visualise` to write out images with boxes displayed.
 
 ## Results
 

diff --git a/models/README.md b/models/README.md
@@ -35,24 +35,6 @@ Trained on [MSCoco](http://cocodataset.org/#download) `train 17`
 
 `yolo3_darknet53_coco`
 
-#### GCV3 (0006 Alternative)
-[**Download**](http://hf.id.au/models/VidDet/GCV3.tar.gz)
-
-Uses MobileNet1.0 instead of DarkNet53
-
-Trained on [PascalVOC](http://host.robots.ox.ac.uk/pascal/VOC/voc2012/index.html#devkit) `trainval 07+12`
-
-`yolo3_mobilenet1.0_voc`
-
-#### GCV4 (0008 Alternative)
-[**Download**](http://hf.id.au/models/VidDet/GCV4.tar.gz)
-
-Uses MobileNet1.0 instead of DarkNet53
-
-Trained on [MSCoco](http://cocodataset.org/#download) `train 17`
-
-`yolo3_mobilenet1.0_coco`
-
 
 ## Our Models
 Our models, log files, and evaluation results are available for download
@@ -64,23 +46,23 @@ by clicking on each model ID below.
 Trained on [PascalVOC](http://host.robots.ox.ac.uk/pascal/VOC/voc2012/index.html#devkit) `trainval 07+12`
 
 ```
-python train_yolov3.py --dataset voc --gpus 0,1,2,3 --save_prefix 0001 --num_workers 16 --warmup_lr 0.0001 --warmup_epochs 3 --syncbn
+python train_yolov3.py --dataset voc --gpus 0,1,2,3 --save_prefix 0001 --num_workers 16 --warmup_epochs 4 --syncbn
 ```
 #### 0002
 [**Download (SOON)**]()
 
 Trained on [ImageNetDET](http://image-net.org/challenges/LSVRC/2017/download-images-1p39.php) `train_nonempty`
 
 ```
-python train_yolov3.py --dataset det --gpus 0,1,2,3 --save_prefix 0002 --num_workers 16 --warmup_lr 0.0001 --warmup_epochs 3 --syncbn
+python train_yolov3.py --dataset det --gpus 0,1,2,3 --save_prefix 0002 --num_workers 16 --warmup_epochs 3 --epochs 140 --lr_decay_epoch 100,120 --syncbn
 ```
 #### 0003
 [**Download**](http://hf.id.au/models/VidDet/0003.tar.gz)
 
 Trained on [MSCoco](http://cocodataset.org/#download) `train 17`
 
 ```
-python train_yolov3.py --dataset coco --gpus 0,1,2,3 --save_prefix 0003 --num_workers 16 --warmup_lr 0.0001 --warmup_epochs 3 --syncbn
+python train_yolov3.py --dataset coco --gpus 0,1,2,3 --save_prefix 0003 --num_workers 16 --warmup_epochs 3 --epochs 280 --lr_decay_epoch2 20,250 --syncbn
 ```
 
 #### 0004
@@ -89,50 +71,9 @@ python train_yolov3.py --dataset coco --gpus 0,1,2,3 --save_prefix 0003 --num_wo
 Trained on [ImageNetVID](http://bvisionweb1.cs.unc.edu/ILSVRC2017/download-videos-1p39.php) `train17_ne_0.04`
 
 ```
-python train_yolov3.py --dataset vid --gpus 0,1,2,3 --save_prefix 0004 --num_workers 16 --warmup_lr 0.0001 --warmup_epochs 3 --syncbn --every 25
-```
-
-#### 0006
-[**Download**](http://hf.id.au/models/VidDet/0006.tar.gz)
-
-Uses MobileNet1.0 instead of DarkNet53
-
-Trained on [PascalVOC](http://host.robots.ox.ac.uk/pascal/VOC/voc2012/index.html#devkit) `trainval 07+12`
-
-```
-python train_yolov3.py --network mobilenet1.0 --dataset voc --gpus 0,1,2,3 --save_prefix 0001 --num_workers 16 --warmup_lr 0.0001 --warmup_epochs 3 --syncbn
-```
-#### 0007
-[**Download (SOON)**]()
-
-Uses MobileNet1.0 instead of DarkNet53
-
-Trained on [ImageNetDET](http://image-net.org/challenges/LSVRC/2017/download-images-1p39.php) `train_nonempty`
-
-```
-python train_yolov3.py --network mobilenet1.0 --dataset det --gpus 0,1,2,3 --save_prefix 0002 --num_workers 16 --warmup_lr 0.0001 --warmup_epochs 3 --syncbn
+python train_yolov3.py --dataset vid --gpus 0,1,2,3 --save_prefix 0004 --num_workers 16 --warmup_epochs 3 --epochs 280 --lr_decay_epoch 220,250 --every 25 --syncbn 
 ```
-#### 0008
-[**Download (SOON)**]()
-
-Uses MobileNet1.0 instead of DarkNet53
 
-Trained on [MSCoco](http://cocodataset.org/#download) `train 17`
-
-```
-python train_yolov3.py --network mobilenet1.0 --dataset coco --gpus 0,1,2,3 --save_prefix 0003 --num_workers 16 --warmup_lr 0.0001 --warmup_epochs 3 --syncbn
-```
-
-#### 0009
-[**Download (SOON)**]()
-
-Uses MobileNet1.0 instead of DarkNet53
-
-Trained on [ImageNetVID](http://bvisionweb1.cs.unc.edu/ILSVRC2017/download-videos-1p39.php) `train17_ne_0.04`
-
-```
-python train_yolov3.py --network mobilenet1_0 --dataset vid --gpus 0,1,2,3 --save_prefix 0004 --num_workers 16 --warmup_lr 0.0001 --warmup_epochs 3 --syncbn --every 25
-```
 
 ## Results
 Evaluated with `voc` and `coco` metrics. Box Area's - **S**mall `<32`,
@@ -142,7 +83,6 @@ Evaluated with `voc` and `coco` metrics. Box Area's - **S**mall `<32`,
 |--------|------------|-----------|------------------|---------------------|-----------|------------------|----------------|----------------|----------------|
 | `0001` | VOC `trainval 07+12` | VOC `test 07` | .835 | .463 | .733 | .510 | .118 | .317 | .559 |
 | `GCV1` | VOC `trainval 07+12` | VOC `test 07` | .836 | .462 | .735 | .500 | .113 | .304 | .564 |
-| `0006` | VOC `trainval 07+12` | VOC `test 07` | .751 | .356 | .656 | .346 | .095 | .205 | .438 |
 | `GCV3` | VOC `trainval 07+12` | VOC `test 07` | .779 | .396 | .677 | .418 | .104 | .245 | .486 |
 | `0003` | COCO `train 17` | COCO `val 17` | .525 | .288 | .515 | .296 | .136 | .306 | .427 |
 | `GCV2` | COCO `train 17` | COCO `val 17` | .579 | .360 | .571 | .387 | .173 | .387 | .522 |