Gitee
commit
2e40ac6465
|
|
@ -82,7 +82,8 @@ Dataset used: [COCO2017](<http://images.cocodataset.org/>)
|
|||
|
||||
# [Quick Start](#contents)
|
||||
|
||||
After installing MindSpore via the official website, you can start training and evaluation on Ascend as follows:
|
||||
After installing MindSpore via the official website, you can start training and evaluation as follows:
|
||||
- runing on Ascend
|
||||
|
||||
```
|
||||
# distributed training on Ascend
|
||||
|
|
@ -91,6 +92,14 @@ sh run_distribute_train.sh [DEVICE_NUM] [EPOCH_SIZE] [LR] [DATASET] [RANK_TABLE_
|
|||
# run eval on Ascend
|
||||
sh run_eval.sh [DATASET] [CHECKPOINT_PATH] [DEVICE_ID]
|
||||
```
|
||||
- runing on GPU
|
||||
```
|
||||
# distributed training on GPU
|
||||
sh run_distribute_train_gpu.sh [DEVICE_NUM] [EPOCH_SIZE] [LR] [DATASET]
|
||||
|
||||
# run eval on GPU
|
||||
sh run_eval_gpu.sh [DATASET] [CHECKPOINT_PATH] [DEVICE_ID]
|
||||
```
|
||||
|
||||
# [Script Description](#contents)
|
||||
|
||||
|
|
@ -100,21 +109,24 @@ sh run_eval.sh [DATASET] [CHECKPOINT_PATH] [DEVICE_ID]
|
|||
.
|
||||
└─ cv
|
||||
└─ ssd
|
||||
├─ README.md ## descriptions about SSD
|
||||
├─ README.md ## descriptions about SSD
|
||||
├─ scripts
|
||||
└─ run_distribute_train.sh ## shell script for distributed on ascend
|
||||
└─ run_eval.sh ## shell script for eval on ascend
|
||||
├─ run_distribute_train.sh ## shell script for distributed on ascend
|
||||
├─ run_distribute_train_gpu.sh ## shell script for distributed on gpu
|
||||
├─ run_eval.sh ## shell script for eval on ascend
|
||||
└─ run_eval_gpu.sh ## shell script for eval on gpu
|
||||
├─ src
|
||||
├─ __init__.py ## init file
|
||||
├─ box_util.py ## bbox utils
|
||||
├─ coco_eval.py ## coco metrics utils
|
||||
├─ config.py ## total config
|
||||
├─ dataset.py ## create dataset and process dataset
|
||||
├─ init_params.py ## parameters utils
|
||||
├─ lr_schedule.py ## learning ratio generator
|
||||
└─ ssd.py ## ssd architecture
|
||||
├─ eval.py ## eval scripts
|
||||
└─ train.py ## train scripts
|
||||
├─ __init__.py ## init file
|
||||
├─ box_util.py ## bbox utils
|
||||
├─ coco_eval.py ## coco metrics utils
|
||||
├─ config.py ## total config
|
||||
├─ dataset.py ## create dataset and process dataset
|
||||
├─ init_params.py ## parameters utils
|
||||
├─ lr_schedule.py ## learning ratio generator
|
||||
└─ ssd.py ## ssd architecture
|
||||
├─ eval.py ## eval scripts
|
||||
├─ train.py ## train scripts
|
||||
└─ mindspore_hub_conf.py ## mindspore hub interface
|
||||
```
|
||||
|
||||
## [Script Parameters](#contents)
|
||||
|
|
@ -145,10 +157,9 @@ sh run_eval.sh [DATASET] [CHECKPOINT_PATH] [DEVICE_ID]
|
|||
|
||||
## [Training Process](#contents)
|
||||
|
||||
### Training on Ascend
|
||||
|
||||
To train the model, run `train.py`. If the `mindrecord_dir` is empty, it will generate [mindrecord](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/convert_dataset.html) files by `coco_root`(coco dataset) or `iamge_dir` and `anno_path`(own dataset). **Note if mindrecord_dir isn't empty, it will use mindrecord_dir instead of raw images.**
|
||||
|
||||
### Training on Ascend
|
||||
|
||||
- Distribute mode
|
||||
|
||||
|
|
@ -183,6 +194,34 @@ epoch: 500 step: 458, loss is 0.5548882
|
|||
epoch time: 39064.8467540741, per step time: 85.29442522723602
|
||||
```
|
||||
|
||||
### Training on GPU
|
||||
|
||||
- Distribute mode
|
||||
|
||||
```
|
||||
sh run_distribute_train_gpu.sh [DEVICE_NUM] [EPOCH_SIZE] [LR] [DATASET] [PRE_TRAINED](optional) [PRE_TRAINED_EPOCH_SIZE](optional)
|
||||
```
|
||||
We need five or seven parameters for this scripts.
|
||||
- `DEVICE_NUM`: the device number for distributed train.
|
||||
- `EPOCH_NUM`: epoch num for distributed train.
|
||||
- `LR`: learning rate init value for distributed train.
|
||||
- `DATASET`：the dataset mode for distributed train.
|
||||
- `PRE_TRAINED :` the path of pretrained checkpoint file, it is better to use absolute path.
|
||||
- `PRE_TRAINED_EPOCH_SIZE :` the epoch num of pretrained.
|
||||
|
||||
Training result will be stored in the current path, whose folder name is "LOG". Under this, you can find checkpoint files together with result like the followings in log
|
||||
|
||||
```
|
||||
epoch: 1 step: 1, loss is 420.11783
|
||||
epoch: 1 step: 2, loss is 434.11032
|
||||
epoch: 1 step: 3, loss is 476.802
|
||||
...
|
||||
epoch: 1 step: 458, loss is 3.1283689
|
||||
epoch time: 150753.701, per step time: 329.157
|
||||
...
|
||||
|
||||
```
|
||||
|
||||
## [Evaluation Process](#contents)
|
||||
|
||||
### Evaluation on Ascend
|
||||
|
|
@ -218,41 +257,73 @@ Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.697
|
|||
mAP: 0.23808886505483504
|
||||
```
|
||||
|
||||
### Evaluation on GPU
|
||||
|
||||
```
|
||||
sh run_eval_gpu.sh [DATASET] [CHECKPOINT_PATH] [DEVICE_ID]
|
||||
```
|
||||
We need two parameters for this scripts.
|
||||
- `DATASET`：the dataset mode of evaluation dataset.
|
||||
- `CHECKPOINT_PATH`: the absolute path for checkpoint file.
|
||||
- `DEVICE_ID`: the device id for eval.
|
||||
|
||||
> checkpoint can be produced in training process.
|
||||
|
||||
Inference result will be stored in the example path, whose folder name begins with "eval". Under this, you can find result like the followings in log.
|
||||
|
||||
```
|
||||
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.224
|
||||
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.375
|
||||
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.228
|
||||
Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.034
|
||||
Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.189
|
||||
Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.407
|
||||
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.243
|
||||
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.382
|
||||
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.417
|
||||
Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.120
|
||||
Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.425
|
||||
Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.686
|
||||
|
||||
========================================
|
||||
|
||||
mAP: 0.2244936111705981
|
||||
```
|
||||
|
||||
# [Model Description](#contents)
|
||||
## [Performance](#contents)
|
||||
|
||||
### Evaluation Performance
|
||||
|
||||
| Parameters | Ascend |
|
||||
| -------------------------- | -------------------------------------------------------------|
|
||||
| Model Version | SSD V1 |
|
||||
| Resource | Ascend 910 ；CPU 2.60GHz，56cores；Memory，314G |
|
||||
| uploaded Date | 06/01/2020 (month/day/year) |
|
||||
| MindSpore Version | 0.3.0-alpha |
|
||||
| Dataset | COCO2017 |
|
||||
| Training Parameters | epoch = 500, batch_size = 32 |
|
||||
| Optimizer | Momentum |
|
||||
| Loss Function | Sigmoid Cross Entropy,SmoothL1Loss |
|
||||
| Speed | 8pcs: 90ms/step |
|
||||
| Total time | 8pcs: 4.81hours |
|
||||
| Parameters (M) | 34 |
|
||||
| Scripts | https://gitee.com/mindspore/mindspore/tree/master/model_zoo/official/cv/ssd |
|
||||
| Parameters | Ascend | GPU |
|
||||
| -------------------------- | -------------------------------------------------------------| -------------------------------------------------------------|
|
||||
| Model Version | SSD V1 | SSD V1 |
|
||||
| Resource | Ascend 910 ；CPU 2.60GHz，56cores；Memory，314G | NV SMX2 V100-16G |
|
||||
| uploaded Date | 06/01/2020 (month/day/year) | 09/24/2020 (month/day/year) |
|
||||
| MindSpore Version | 0.3.0-alpha | 1.0.0 |
|
||||
| Dataset | COCO2017 | COCO2017 |
|
||||
| Training Parameters | epoch = 500, batch_size = 32 | epoch = 800, batch_size = 32 |
|
||||
| Optimizer | Momentum | Momentum |
|
||||
| Loss Function | Sigmoid Cross Entropy,SmoothL1Loss | Sigmoid Cross Entropy,SmoothL1Loss |
|
||||
| Speed | 8pcs: 90ms/step | 8pcs: 121ms/step |
|
||||
| Total time | 8pcs: 4.81hours | 8pcs: 12.31hours |
|
||||
| Parameters (M) | 34 | 34 |
|
||||
| Scripts | https://gitee.com/mindspore/mindspore/tree/master/model_zoo/official/cv/ssd | https://gitee.com/mindspore/mindspore/tree/master/model_zoo/official/cv/ssd |
|
||||
|
||||
|
||||
### Inference Performance
|
||||
|
||||
| Parameters | Ascend |
|
||||
| ------------------- | ----------------------------|
|
||||
| Model Version | SSD V1 |
|
||||
| Resource | Ascend 910 |
|
||||
| Uploaded Date | 06/01/2020 (month/day/year) |
|
||||
| MindSpore Version | 0.3.0-alpha |
|
||||
| Dataset | COCO2017 |
|
||||
| batch_size | 1 |
|
||||
| outputs | mAP |
|
||||
| Accuracy | IoU=0.50: 23.8% |
|
||||
| Model for inference | 34M(.ckpt file) |
|
||||
| Parameters | Ascend | GPU |
|
||||
| ------------------- | ----------------------------| ----------------------------|
|
||||
| Model Version | SSD V1 | SSD V1 |
|
||||
| Resource | Ascend 910 | GPU |
|
||||
| Uploaded Date | 06/01/2020 (month/day/year) | 09/24/2020 (month/day/year) |
|
||||
| MindSpore Version | 0.3.0-alpha | 1.0.0 |
|
||||
| Dataset | COCO2017 | COCO2017 |
|
||||
| batch_size | 1 | 1 |
|
||||
| outputs | mAP | mAP |
|
||||
| Accuracy | IoU=0.50: 23.8% | IoU=0.50: 22.4% |
|
||||
| Model for inference | 34M(.ckpt file) | 34M(.ckpt file) |
|
||||
|
||||
# [Description of Random Situation](#contents)
|
||||
|
||||
|
|
|
|||
|
|
@ -71,9 +71,11 @@ if __name__ == '__main__':
|
|||
parser.add_argument("--device_id", type=int, default=0, help="Device id, default is 0.")
|
||||
parser.add_argument("--dataset", type=str, default="coco", help="Dataset, default is coco.")
|
||||
parser.add_argument("--checkpoint_path", type=str, required=True, help="Checkpoint file path.")
|
||||
parser.add_argument("--run_platform", type=str, default="Ascend", choices=("Ascend", "GPU"),
|
||||
help="run platform, only support Ascend and GPU.")
|
||||
args_opt = parser.parse_args()
|
||||
|
||||
context.set_context(mode=context.GRAPH_MODE, device_target="Ascend", device_id=args_opt.device_id)
|
||||
context.set_context(mode=context.GRAPH_MODE, device_target=args_opt.run_platform, device_id=args_opt.device_id)
|
||||
|
||||
prefix = "ssd_eval.mindrecord"
|
||||
mindrecord_dir = config.mindrecord_dir
|
||||
|
|
|
|||
|
|
@ -0,0 +1,77 @@
|
|||
#!/bin/bash
|
||||
# Copyright 2020 Huawei Technologies Co., Ltd
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
# ============================================================================
|
||||
|
||||
echo "=============================================================================================================="
|
||||
echo "Please run the scipt as: "
|
||||
echo "sh run_distribute_train_gpu.sh DEVICE_NUM EPOCH_SIZE LR DATASET PRE_TRAINED PRE_TRAINED_EPOCH_SIZE"
|
||||
echo "for example: sh run_distribute_train_gpu.sh 8 500 0.2 coco /opt/ssd-300.ckpt(optional) 200(optional)"
|
||||
echo "It is better to use absolute path."
|
||||
echo "================================================================================================================="
|
||||
|
||||
if [ $# != 4 ] && [ $# != 6 ]
|
||||
then
|
||||
echo "Usage: sh run_distribute_train_gpu.sh [DEVICE_NUM] [EPOCH_SIZE] [LR] [DATASET] \
|
||||
[PRE_TRAINED](optional) [PRE_TRAINED_EPOCH_SIZE](optional)"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
# Before start distribute train, first create mindrecord files.
|
||||
BASE_PATH=$(cd "`dirname $0`" || exit; pwd)
|
||||
cd $BASE_PATH/../ || exit
|
||||
python train.py --only_create_dataset=True --run_platform="GPU"
|
||||
|
||||
echo "After running the scipt, the network runs in the background. The log will be generated in LOG/log.txt"
|
||||
|
||||
export RANK_SIZE=$1
|
||||
EPOCH_SIZE=$2
|
||||
LR=$3
|
||||
DATASET=$4
|
||||
PRE_TRAINED=$5
|
||||
PRE_TRAINED_EPOCH_SIZE=$6
|
||||
|
||||
rm -rf LOG
|
||||
mkdir ./LOG
|
||||
cp ./*.py ./LOG
|
||||
cp -r ./src ./LOG
|
||||
cd ./LOG || exit
|
||||
|
||||
if [ $# == 4 ]
|
||||
then
|
||||
mpirun -allow-run-as-root -n $RANK_SIZE --output-filename log_output --merge-stderr-to-stdout \
|
||||
python train.py \
|
||||
--distribute=True \
|
||||
--lr=$LR \
|
||||
--dataset=$DATASET \
|
||||
--device_num=$RANK_SIZE \
|
||||
--loss_scale=1 \
|
||||
--run_platform="GPU" \
|
||||
--epoch_size=$EPOCH_SIZE > log.txt 2>&1 &
|
||||
fi
|
||||
|
||||
if [ $# == 6 ]
|
||||
then
|
||||
mpirun -allow-run-as-root -n $RANK_SIZE --output-filename log_output --merge-stderr-to-stdout \
|
||||
python train.py \
|
||||
--distribute=True \
|
||||
--lr=$LR \
|
||||
--dataset=$DATASET \
|
||||
--device_num=$RANK_SIZE \
|
||||
--pre_trained=$PRE_TRAINED \
|
||||
--pre_trained_epoch_size=$PRE_TRAINED_EPOCH_SIZE \
|
||||
--loss_scale=1 \
|
||||
--run_platform="GPU" \
|
||||
--epoch_size=$EPOCH_SIZE > log.txt 2>&1 &
|
||||
fi
|
||||
|
|
@ -0,0 +1,66 @@
|
|||
#!/bin/bash
|
||||
# Copyright 2020 Huawei Technologies Co., Ltd
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
# ============================================================================
|
||||
|
||||
if [ $# != 3 ]
|
||||
then
|
||||
echo "Usage: sh run_eval_gpu.sh [DATASET] [CHECKPOINT_PATH] [DEVICE_ID]"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
get_real_path(){
|
||||
if [ "${1:0:1}" == "/" ]; then
|
||||
echo "$1"
|
||||
else
|
||||
echo "$(realpath -m $PWD/$1)"
|
||||
fi
|
||||
}
|
||||
|
||||
DATASET=$1
|
||||
CHECKPOINT_PATH=$(get_real_path $2)
|
||||
echo $DATASET
|
||||
echo $CHECKPOINT_PATH
|
||||
|
||||
if [ ! -f $CHECKPOINT_PATH ]
|
||||
then
|
||||
echo "error: CHECKPOINT_PATH=$PATH2 is not a file"
|
||||
exit 1
|
||||
fi
|
||||
|
||||
export DEVICE_NUM=1
|
||||
export DEVICE_ID=$3
|
||||
export RANK_SIZE=$DEVICE_NUM
|
||||
export RANK_ID=0
|
||||
|
||||
BASE_PATH=$(cd "`dirname $0`" || exit; pwd)
|
||||
cd $BASE_PATH/../ || exit
|
||||
|
||||
if [ -d "eval$3" ];
|
||||
then
|
||||
rm -rf ./eval$3
|
||||
fi
|
||||
|
||||
mkdir ./eval$3
|
||||
cp ./*.py ./eval$3
|
||||
cp -r ./src ./eval$3
|
||||
cd ./eval$3 || exit
|
||||
env > env.log
|
||||
echo "start infering for device $DEVICE_ID"
|
||||
python eval.py \
|
||||
--dataset=$DATASET \
|
||||
--checkpoint_path=$CHECKPOINT_PATH \
|
||||
--run_platform="GPU" \
|
||||
--device_id=$3 > log.txt 2>&1 &
|
||||
cd ..
|
||||
|
|
@ -250,6 +250,8 @@ class SSD300(nn.Cell):
|
|||
pred_loc, pred_label = self.multi_box(multi_feature)
|
||||
if not self.is_training:
|
||||
pred_label = self.activation(pred_label)
|
||||
pred_loc = F.cast(pred_loc, mstype.float32)
|
||||
pred_label = F.cast(pred_label, mstype.float32)
|
||||
return pred_loc, pred_label
|
||||
|
||||
|
||||
|
|
|
|||
|
|
@ -20,12 +20,12 @@ import argparse
|
|||
import ast
|
||||
import mindspore.nn as nn
|
||||
from mindspore import context, Tensor
|
||||
from mindspore.communication.management import init
|
||||
from mindspore.communication.management import init, get_rank
|
||||
from mindspore.train.callback import CheckpointConfig, ModelCheckpoint, LossMonitor, TimeMonitor
|
||||
from mindspore.train import Model
|
||||
from mindspore.context import ParallelMode
|
||||
from mindspore.train.serialization import load_checkpoint, load_param_into_net
|
||||
from mindspore.common import set_seed
|
||||
from mindspore.common import set_seed, dtype
|
||||
from src.ssd import SSD300, SSDWithLossCell, TrainingWrapper, ssd_mobilenet_v2
|
||||
from src.config import config
|
||||
from src.dataset import create_ssd_dataset, data_to_mindrecord_byte_image, voc_data_to_mindrecord
|
||||
|
|
@ -53,20 +53,36 @@ def main():
|
|||
parser.add_argument("--loss_scale", type=int, default=1024, help="Loss scale, default is 1024.")
|
||||
parser.add_argument("--filter_weight", type=ast.literal_eval, default=False,
|
||||
help="Filter weight parameters, default is False.")
|
||||
parser.add_argument("--run_platform", type=str, default="Ascend", choices=("Ascend", "GPU"),
|
||||
help="run platform, only support Ascend and GPU.")
|
||||
args_opt = parser.parse_args()
|
||||
|
||||
context.set_context(mode=context.GRAPH_MODE, device_target="Ascend", device_id=args_opt.device_id)
|
||||
|
||||
if args_opt.distribute:
|
||||
device_num = args_opt.device_num
|
||||
context.reset_auto_parallel_context()
|
||||
context.set_auto_parallel_context(parallel_mode=ParallelMode.DATA_PARALLEL, gradients_mean=True,
|
||||
device_num=device_num)
|
||||
if args_opt.run_platform == "Ascend":
|
||||
context.set_context(mode=context.GRAPH_MODE, device_target="Ascend", device_id=args_opt.device_id)
|
||||
if args_opt.distribute:
|
||||
device_num = args_opt.device_num
|
||||
context.reset_auto_parallel_context()
|
||||
context.set_auto_parallel_context(parallel_mode=ParallelMode.DATA_PARALLEL, gradients_mean=True,
|
||||
device_num=device_num)
|
||||
init()
|
||||
rank = args_opt.device_id % device_num
|
||||
else:
|
||||
rank = 0
|
||||
device_num = 1
|
||||
elif args_opt.run_platform == "GPU":
|
||||
context.set_context(mode=context.GRAPH_MODE, device_target="GPU", device_id=args_opt.device_id)
|
||||
init()
|
||||
rank = args_opt.device_id % device_num
|
||||
if args_opt.distribute:
|
||||
device_num = args_opt.device_num
|
||||
context.reset_auto_parallel_context()
|
||||
context.set_auto_parallel_context(parallel_mode=ParallelMode.DATA_PARALLEL, gradients_mean=True,
|
||||
device_num=device_num)
|
||||
rank = get_rank()
|
||||
else:
|
||||
rank = 0
|
||||
device_num = 1
|
||||
else:
|
||||
rank = 0
|
||||
device_num = 1
|
||||
raise ValueError("Unsupported platform.")
|
||||
|
||||
print("Start create dataset!")
|
||||
|
||||
|
|
@ -113,6 +129,8 @@ def main():
|
|||
|
||||
backbone = ssd_mobilenet_v2()
|
||||
ssd = SSD300(backbone=backbone, config=config)
|
||||
if args_opt.run_platform == "GPU":
|
||||
ssd.to_float(dtype.float16)
|
||||
net = SSDWithLossCell(ssd, config)
|
||||
init_net_param(net)
|
||||
|
||||
|
|
|
|||
Loading…
Reference in New Issue