Skip to content

Commit b27abeb

Browse files
nvpstrnvpstr
committed
[UNet_medical/TF1&2] Updating for Ampere
1 parent 76a056c commit b27abeb

65 files changed

Lines changed: 781 additions & 882 deletions

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

‎TensorFlow/Segmentation/UNet_Medical/.gitmodules‎

Whitespace-only changes.

‎TensorFlow/Segmentation/UNet_Medical/Dockerfile‎

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,7 +1,8 @@
1-
ARG FROM_IMAGE_NAME=nvcr.io/nvidia/tensorflow:20.01-tf1-py3
1+
ARG FROM_IMAGE_NAME=nvcr.io/nvidia/tensorflow:20.06-tf1-py3
22
FROM ${FROM_IMAGE_NAME}
33

44
ADD . /workspace/unet
55
WORKDIR /workspace/unet
66

7+
RUN pip install git+https://github.com/NVIDIA/dllogger
78
RUN pip install -r requirements.txt

‎TensorFlow/Segmentation/UNet_Medical/README.md‎

Lines changed: 194 additions & 117 deletions
Large diffs are not rendered by default.

‎TensorFlow/Segmentation/UNet_Medical/dllogger/logger.py‎

Lines changed: 0 additions & 163 deletions
This file was deleted.

TensorFlow/Segmentation/UNet_Medical/examples/unet_FP32_1GPU.sh renamed to TensorFlow/Segmentation/UNet_Medical/examples/unet_1GPU.sh

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,4 @@
1-
# Copyright (c) 2019, NVIDIA CORPORATION. All rights reserved.
1+
# Copyright (c) 2020, NVIDIA CORPORATION. All rights reserved.
22
#
33
# Licensed under the Apache License, Version 2.0 (the "License");
44
# you may not use this file except in compliance with the License.
@@ -12,7 +12,7 @@
1212
# See the License for the specific language governing permissions and
1313
# limitations under the License.
1414

15-
# This script launches U-Net run in FP32 on 1 GPU and trains for 40000 iterations with batch_size 1. Usage:
15+
# This script launches U-Net run in FP32 on 1 GPU and trains for 6400 iterations with batch_size 8. Usage:
1616
# bash unet_FP32_1GPU.sh <path to dataset> <path to results directory>
1717

18-
horovodrun -np 1 python main.py --data_dir $1 --model_dir $2 --log_every 100 --max_steps 40000 --batch_size 1 --exec_mode train_and_evaluate --crossvalidation_idx 0 --augment --use_xla --log_dir $2
18+
horovodrun -np 1 python main.py --data_dir $1 --model_dir $2 --log_every 100 --max_steps 6400 --batch_size 8 --exec_mode train_and_evaluate --crossvalidation_idx 0 --augment --xla

TensorFlow/Segmentation/UNet_Medical/examples/unet_FP32_8GPU.sh renamed to TensorFlow/Segmentation/UNet_Medical/examples/unet_8GPU.sh

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,4 @@
1-
# Copyright (c) 2019, NVIDIA CORPORATION. All rights reserved.
1+
# Copyright (c) 2020, NVIDIA CORPORATION. All rights reserved.
22
#
33
# Licensed under the Apache License, Version 2.0 (the "License");
44
# you may not use this file except in compliance with the License.
@@ -12,7 +12,7 @@
1212
# See the License for the specific language governing permissions and
1313
# limitations under the License.
1414

15-
# This script launches U-Net run in FP32 on 8 GPUs and trains for 40000 iterations with batch_size 1. Usage:
15+
# This script launches U-Net run in FP32 on 8 GPUs and trains for 6400 iterations with batch_size 8. Usage:
1616
# bash unet_FP32_8GPU.sh <path to dataset> <path to results directory>
1717

18-
horovodrun -np 8 python main.py --data_dir $1 --model_dir $2 --log_every 100 --max_steps 40000 --batch_size 1 --exec_mode train_and_evaluate --crossvalidation_idx 0 --augment --use_xla --log_dir $2
18+
horovodrun -np 8 python main.py --data_dir $1 --model_dir $2 --log_every 100 --max_steps 6400 --batch_size 8 --exec_mode train_and_evaluate --crossvalidation_idx 0 --augment --xla

TensorFlow/Segmentation/UNet_Medical/examples/unet_INFER_FP32.sh renamed to TensorFlow/Segmentation/UNet_Medical/examples/unet_INFER.sh

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -15,4 +15,4 @@
1515
# This script launches U-Net run in FP32 on 1 GPU for inference batch_size 1. Usage:
1616
# bash unet_INFER_FP32.sh <path to this repository> <path to dataset> <path to results directory>
1717

18-
horovodrun -np 1 python main.py --data_dir $1 --model_dir $2 --batch_size 1 --exec_mode predict --use_xla
18+
horovodrun -np 1 python main.py --data_dir $1 --model_dir $2 --batch_size 1 --exec_mode predict --xla

TensorFlow/Segmentation/UNet_Medical/examples/unet_INFER_BENCHMARK_FP32.sh renamed to TensorFlow/Segmentation/UNet_Medical/examples/unet_INFER_BENCHMARK.sh

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -15,4 +15,4 @@
1515
# This script launches U-Net run in FP32 on 1 GPU for inference benchmarking. Usage:
1616
# bash unet_INFER_BENCHMARK_FP32.sh <path to dataset> <path to results directory> <batch size>
1717

18-
horovodrun -np 1 python main.py --data_dir $1 --model_dir $2 --batch_size $3 --exec_mode predict --benchmark --warmup_steps 200 --max_steps 600 --use_xla
18+
horovodrun -np 1 python main.py --data_dir $1 --model_dir $2 --batch_size $3 --exec_mode predict --benchmark --warmup_steps 200 --max_steps 600 --xla

‎TensorFlow/Segmentation/UNet_Medical/examples/unet_INFER_BENCHMARK_TF-AMP.sh‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -15,4 +15,4 @@
1515
# This script launches U-Net run in FP16 on 1 GPU for inference benchmarking. Usage:
1616
# bash unet_INFER_BENCHMARK_TF-AMP.sh <path to dataset> <path to results directory> <batch size>
1717

18-
horovodrun -np 1 python main.py --data_dir $1 --model_dir $2 --batch_size $3 --exec_mode predict --benchmark --warmup_steps 200 --max_steps 600 --use_xla --use_amp
18+
horovodrun -np 1 python main.py --data_dir $1 --model_dir $2 --batch_size $3 --exec_mode predict --benchmark --warmup_steps 200 --max_steps 600 --xla --amp

‎TensorFlow/Segmentation/UNet_Medical/examples/unet_INFER_BENCHMARK_TF-TRT.sh‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -15,4 +15,4 @@
1515
# This script launches U-Net training in FP32 on 1 GPUs using 2 batch size
1616
# Usage ./unet_INFER_BENCHMARK_FP32.sh <path to this repository> <path to dataset> <path to results directory> <batch size>
1717

18-
python $1/main.py --data_dir $2 --model_dir $3 --batch_size $4 --benchmark --exec_mode predict --augment --warmup_steps 200 --log_every 100 --max_steps 300 --use_xla
18+
python $1/main.py --data_dir $2 --model_dir $3 --batch_size $4 --benchmark --exec_mode predict --augment --warmup_steps 200 --log_every 100 --max_steps 300 --xla

0 commit comments

Comments
 (0)