maxmustermann/DeepLearningExamples
1
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity

Fix README

Browse source
This commit is contained in:
sharatht 2019-05-23 12:21:00 -07:00
parent b1ae8dd47c
commit f2bc6ae8ab
2 changed files with 6 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
PyTorch/Segmentation/MaskRCNN/README.md
View file

@ -103,7 +103,7 @@ To train your model using mixed precision with tensor cores or using FP32, perfo
### 1. Clone the repository.
```
git clone https://github.com/NVIDIA/DeepLearningExamples.git
cd DeepLearningExamples/PyTorch/Segmentation/MaskRCNN
cd DeepLearningExamples/PyTorch/Segmentation
```
### 2. Download and preprocess the dataset.
@ -111,7 +111,6 @@ This repository provides scripts to download and extract the COCO 2014 dataset.
To download, verify, and extract the COCO dataset, use the following scripts:
```
cd Detectron_PyT
./download_dataset.sh <data/dir>
```
By default, the data is organized into the following structure:
@ -128,6 +127,7 @@ By default, the data is organized into the following structure:
### 3. Build the Mask R-CNN PyTorch NGC container.
```
cd MaskRCNN/
bash scripts/docker/build.sh
```

7
PyTorch/Segmentation/MaskRCNN/pytorch/configs/e2e_mask_rcnn_R_50_FPN_1x.yaml
View file

@ -11,6 +11,7 @@ MODEL:
PRE_NMS_TOP_N_TEST: 1000
POST_NMS_TOP_N_TEST: 1000
FPN_POST_NMS_TOP_N_TEST: 1000
FPN_POST_NMS_TOP_N_TRAIN: 4000
ROI_HEADS:
USE_FPN: True
ROI_BOX_HEAD:
@ -36,8 +37,8 @@ DATALOADER:
SOLVER:
BASE_LR: 0.04
WEIGHT_DECAY: 0.0001
STEPS: (30000, 40000)
MAX_ITER: 45000
STEPS: (36000, 48000)
MAX_ITER: 50000
IMS_PER_BATCH: 32
TEST:
IMS_PER_BATCH: 32
IMS_PER_BATCH: 8