From 4011c36abc84a22531753241839fae4487e23074 Mon Sep 17 00:00:00 2001 From: lvmingfu Date: Wed, 14 Jul 2021 10:35:49 +0800 Subject: [PATCH] modify urls related docs repository for master_mindspore --- model_zoo/official/cv/FCN8s/README.md | 6 +++--- model_zoo/official/cv/cnnctc/README.md | 2 +- model_zoo/official/cv/cnnctc/README_CN.md | 4 ++-- model_zoo/official/cv/crnn/README.md | 2 +- model_zoo/official/cv/crnn_seq2seq_ocr/README.md | 2 +- model_zoo/official/cv/cspdarknet53/README.md | 4 ++-- model_zoo/official/cv/ctpn/README.md | 2 +- model_zoo/official/cv/deeplabv3/README.md | 6 +++--- model_zoo/official/cv/deeplabv3/README_CN.md | 2 +- model_zoo/official/cv/deeptext/README.md | 2 +- model_zoo/official/cv/densenet/README.md | 2 +- model_zoo/official/cv/densenet/README_CN.md | 2 +- model_zoo/official/cv/dpn/README.md | 2 +- model_zoo/official/cv/googlenet/README.md | 2 +- model_zoo/official/cv/googlenet/README_CN.md | 8 ++++---- model_zoo/official/cv/inceptionv3/README.md | 4 ++-- model_zoo/official/cv/inceptionv3/README_CN.md | 4 ++-- model_zoo/official/cv/inceptionv4/README.md | 4 ++-- model_zoo/official/cv/lenet_quant/Readme_CN.md | 6 +++--- model_zoo/official/cv/maskrcnn/README.md | 2 +- model_zoo/official/cv/maskrcnn/README_CN.md | 2 +- model_zoo/official/cv/maskrcnn_mobilenetv1/README.md | 2 +- model_zoo/official/cv/mobilenetv1/README.md | 2 +- model_zoo/official/cv/mobilenetv2/README.md | 2 +- model_zoo/official/cv/mobilenetv2/README_CN.md | 2 +- model_zoo/official/cv/mobilenetv2_quant/README_CN.md | 2 +- model_zoo/official/cv/mobilenetv2_quant/Readme.md | 2 +- model_zoo/official/cv/nasnet/README_CN.md | 4 ++-- model_zoo/official/cv/openpose/README.md | 2 +- model_zoo/official/cv/resnet/README.md | 4 ++-- model_zoo/official/cv/resnet/README_CN.md | 2 +- model_zoo/official/cv/resnet50_quant/README.md | 2 +- model_zoo/official/cv/resnet50_quant/README_CN.md | 6 +++--- model_zoo/official/cv/resnext/README.md | 2 +- model_zoo/official/cv/resnext/README_CN.md | 2 +- model_zoo/official/cv/retinanet/README_CN.md | 8 ++++---- model_zoo/official/cv/simple_pose/README.md | 2 +- model_zoo/official/cv/squeezenet/README.md | 2 +- model_zoo/official/cv/ssd/README.md | 2 +- model_zoo/official/cv/ssd/README_CN.md | 2 +- model_zoo/official/cv/tinydarknet/README_CN.md | 8 ++++---- model_zoo/official/cv/unet3d/README.md | 2 +- model_zoo/official/cv/vgg16/README.md | 4 ++-- model_zoo/official/cv/vgg16/README_CN.md | 4 ++-- model_zoo/official/cv/warpctc/README.md | 2 +- model_zoo/official/cv/warpctc/README_CN.md | 2 +- model_zoo/official/cv/xception/README.md | 4 ++-- model_zoo/official/cv/yolov3_resnet18/README.md | 4 ++-- model_zoo/official/cv/yolov3_resnet18/README_CN.md | 4 ++-- model_zoo/official/nlp/bert/README.md | 2 +- model_zoo/official/nlp/cpm/README.md | 6 +++--- model_zoo/official/nlp/cpm/README_CN.md | 2 +- model_zoo/official/nlp/emotect/README_CN.md | 4 ++-- model_zoo/official/nlp/pangu_alpha/README.md | 6 +++--- model_zoo/official/nlp/transformer/README.md | 2 +- model_zoo/official/nlp/transformer/README_CN.md | 2 +- model_zoo/official/recommend/ncf/README.md | 4 ++-- model_zoo/research/audio/fcn-4/README.md | 2 +- model_zoo/research/cv/AttGAN/README_CN.md | 4 ++-- model_zoo/research/cv/GENet_Res50/README_CN.md | 6 +++--- model_zoo/research/cv/ManiDP/Readme.md | 4 ++-- model_zoo/research/cv/SE-Net/README.md | 2 +- model_zoo/research/cv/arcface/README.md | 2 +- model_zoo/research/cv/deeplabv3plus/README_CN.md | 2 +- model_zoo/research/cv/glore_res200/README_CN.md | 2 +- model_zoo/research/cv/glore_res50/README.md | 4 ++-- model_zoo/research/cv/hardnet/README_CN.md | 10 +++++----- model_zoo/research/cv/ntsnet/README.md | 2 +- model_zoo/research/cv/renas/Readme.md | 4 ++-- model_zoo/research/cv/retinanet_resnet101/README_CN.md | 8 ++++---- model_zoo/research/cv/retinanet_resnet152/README_CN.md | 8 ++++---- model_zoo/research/cv/squeezenet/README.md | 2 +- model_zoo/research/cv/ssd_ghostnet/README.md | 2 +- model_zoo/research/cv/ssd_mobilenetV2/README.md | 2 +- .../research/cv/ssd_mobilenetV2_FPNlite/README.md | 2 +- model_zoo/research/cv/ssd_resnet50/README.md | 2 +- model_zoo/research/cv/ssd_resnet50/README_CN.md | 2 +- model_zoo/research/cv/vgg19/README_CN.md | 2 +- model_zoo/research/cv/wgan/README_CN.md | 4 ++-- model_zoo/research/hpc/pinns/README.md | 2 +- model_zoo/research/hpc/pinns/README_CN.md | 6 +++--- model_zoo/research/hpc/sponge/README.md | 2 +- 82 files changed, 136 insertions(+), 136 deletions(-) diff --git a/model_zoo/official/cv/FCN8s/README.md b/model_zoo/official/cv/FCN8s/README.md index 183d29b2eab..40cfb0dbcc9 100755 --- a/model_zoo/official/cv/FCN8s/README.md +++ b/model_zoo/official/cv/FCN8s/README.md @@ -49,8 +49,8 @@ Dataset used: - 框架 - [MindSpore](https://www.mindspore.cn/install/en) - 如需获取更多信息，请查看如下链接： - - [MindSpore Tutorials](https://www.mindspore.cn/tutorials/en/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore Tutorials](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) # [快速开始](#contents) @@ -434,7 +434,7 @@ python export.py ### 教程 -如果你需要在不同硬件平台（如GPU，Ascend 910 或者 Ascend 310）使用训练好的模型，你可以参考这个 [Link](https://www.mindspore.cn/docs/programming_guide/en/master/multi_platform_inference.html)。以下是一个简单例子的步骤介绍： +如果你需要在不同硬件平台（如GPU，Ascend 910 或者 Ascend 310）使用训练好的模型，你可以参考这个 [Link](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/multi_platform_inference.html)。以下是一个简单例子的步骤介绍： - Running on Ascend diff --git a/model_zoo/official/cv/cnnctc/README.md b/model_zoo/official/cv/cnnctc/README.md index c94a261bceb..e898cd319bb 100644 --- a/model_zoo/official/cv/cnnctc/README.md +++ b/model_zoo/official/cv/cnnctc/README.md @@ -94,7 +94,7 @@ This takes around 75 minutes. ## Mixed Precision -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) diff --git a/model_zoo/official/cv/cnnctc/README_CN.md b/model_zoo/official/cv/cnnctc/README_CN.md index a154a7c9768..d05f8d53943 100644 --- a/model_zoo/official/cv/cnnctc/README_CN.md +++ b/model_zoo/official/cv/cnnctc/README_CN.md @@ -95,7 +95,7 @@ python src/preprocess_dataset.py ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 # 环境要求 @@ -236,7 +236,7 @@ bash scripts/run_distribute_train_ascend.sh [RANK_TABLE_FILE] [PRETRAINED_CKPT(o > 注意: - RANK_TABLE_FILE相关参考资料见[链接](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/distributed_training_ascend.html), 获取device_ip方法详见[链接](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). + RANK_TABLE_FILE相关参考资料见[链接](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/distributed_training_ascend.html), 获取device_ip方法详见[链接](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). ### 训练结果 diff --git a/model_zoo/official/cv/crnn/README.md b/model_zoo/official/cv/crnn/README.md index 00a0ece0f6b..13dadd341a1 100644 --- a/model_zoo/official/cv/crnn/README.md +++ b/model_zoo/official/cv/crnn/README.md @@ -204,7 +204,7 @@ max_text_length": 23, # max number of digits in each ## [Training Process](#contents) -- Set options in `config.py`, including learning rate and other network hyperparameters. Click [MindSpore dataset preparation tutorial](https://www.mindspore.cn/tutorial/training/zh-CN/master/use/data_preparation.html) for more information about dataset. +- Set options in `config.py`, including learning rate and other network hyperparameters. Click [MindSpore dataset preparation tutorial](https://www.mindspore.cn/docs/programming_guide/en/master/dataset_sample.html) for more information about dataset. ### [Training](#contents) diff --git a/model_zoo/official/cv/crnn_seq2seq_ocr/README.md b/model_zoo/official/cv/crnn_seq2seq_ocr/README.md index dbde6e4f1df..aa6148957d9 100755 --- a/model_zoo/official/cv/crnn_seq2seq_ocr/README.md +++ b/model_zoo/official/cv/crnn_seq2seq_ocr/README.md @@ -174,7 +174,7 @@ Parameters for both training and evaluation can be set in config.py. ## [Training Process](#contents) -- Set options in `default_config.yaml`, including learning rate and other network hyperparameters. Click [MindSpore dataset preparation tutorial](https://www.mindspore.cn/tutorial/training/zh-CN/master/use/data_preparation.html) for more information about dataset. +- Set options in `default_config.yaml`, including learning rate and other network hyperparameters. Click [MindSpore dataset preparation tutorial](https://www.mindspore.cn/docs/programming_guide/en/master/dataset_sample.html) for more information about dataset. ### [Training](#contents) diff --git a/model_zoo/official/cv/cspdarknet53/README.md b/model_zoo/official/cv/cspdarknet53/README.md index 1a694a36d6b..88aaeeeb2bb 100644 --- a/model_zoo/official/cv/cspdarknet53/README.md +++ b/model_zoo/official/cv/cspdarknet53/README.md @@ -49,7 +49,7 @@ Dataset used can refer to paper. ## [Mixed Precision(Ascend)](#contents) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. @@ -206,7 +206,7 @@ bash run_distribute_train.sh [RANK_TABLE_FILE] [DATA_DIR] (option)[PATH_CHECKPOI bash run_standalone_train.sh [DEVICE_ID] [DATA_DIR] (option)[PATH_CHECKPOINT] ``` -> Notes: RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/distributed_training_ascend.html), and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). For large models like InceptionV3, it's better to export an external environment variable `export HCCL_CONNECT_TIMEOUT=600` to extend hccl connection checking time from the default 120 seconds to 600 seconds. Otherwise, the connection could be timeout since compiling time increases with the growth of model size. +> Notes: RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/docs/programming_guide/en/master/distributed_training_ascend.html), and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). For large models like InceptionV3, it's better to export an external environment variable `export HCCL_CONNECT_TIMEOUT=600` to extend hccl connection checking time from the default 120 seconds to 600 seconds. Otherwise, the connection could be timeout since compiling time increases with the growth of model size. > > This is processor cores binding operation regarding the `device_num` and total processor numbers. If you are not expect to do it, remove the operations `taskset` in `scripts/run_distribute_train.sh` diff --git a/model_zoo/official/cv/ctpn/README.md b/model_zoo/official/cv/ctpn/README.md index f30236ecb0d..014fe768ef8 100644 --- a/model_zoo/official/cv/ctpn/README.md +++ b/model_zoo/official/cv/ctpn/README.md @@ -195,7 +195,7 @@ imagenet_cfg = edict({ Then you can train it with ImageNet2012. > Notes: -> RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/distributed_training_ascend.html) , and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). For large models like InceptionV4, it's better to export an external environment variable `export HCCL_CONNECT_TIMEOUT=600` to extend hccl connection checking time from the default 120 seconds to 600 seconds. Otherwise, the connection could be timeout since compiling time increases with the growth of model size. +> RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/docs/programming_guide/en/master/distributed_training_ascend.html) , and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). For large models like InceptionV4, it's better to export an external environment variable `export HCCL_CONNECT_TIMEOUT=600` to extend hccl connection checking time from the default 120 seconds to 600 seconds. Otherwise, the connection could be timeout since compiling time increases with the growth of model size. > > This is processor cores binding operation regarding the `device_num` and total processor numbers. If you are not expect to do it, remove the operations `taskset` in `scripts/run_distribute_train.sh` > diff --git a/model_zoo/official/cv/deeplabv3/README.md b/model_zoo/official/cv/deeplabv3/README.md index 67fe84aeeec..6bd78c34abb 100644 --- a/model_zoo/official/cv/deeplabv3/README.md +++ b/model_zoo/official/cv/deeplabv3/README.md @@ -68,7 +68,7 @@ You can also generate the list file automatically by run script: `python get_dat ## Mixed Precision -The [mixed precision](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data types, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data types, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) @@ -78,8 +78,8 @@ For FP16 operators, if the input data type is FP32, the backend of MindSpore wil - Framework - [MindSpore](https://www.mindspore.cn/install/en) - For more information, please check the resources below： - - [MindSpore Tutorials](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) + - [MindSpore Tutorials](https://www.mindspore.cn/tutorials/en/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) - Install python packages in requirements.txt - Generate config json file for 8pcs training diff --git a/model_zoo/official/cv/deeplabv3/README_CN.md b/model_zoo/official/cv/deeplabv3/README_CN.md index 6aaf911bbec..15094b3c8e9 100644 --- a/model_zoo/official/cv/deeplabv3/README_CN.md +++ b/model_zoo/official/cv/deeplabv3/README_CN.md @@ -93,7 +93,7 @@ python convert_resnet101.py ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 # 环境要求 diff --git a/model_zoo/official/cv/deeptext/README.md b/model_zoo/official/cv/deeptext/README.md index 5bc95a67658..16beefbe809 100644 --- a/model_zoo/official/cv/deeptext/README.md +++ b/model_zoo/official/cv/deeptext/README.md @@ -116,7 +116,7 @@ sh run_eval_ascend.sh [IMGS_PATH] [ANNOS_PATH] [CHECKPOINT_PATH] [COCO_TEXT_PARS ``` > Notes: -> RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/distributed_training_ascend.html) , and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). For large models like InceptionV4, it's better to export an external environment variable `export HCCL_CONNECT_TIMEOUT=600` to extend hccl connection checking time from the default 120 seconds to 600 seconds. Otherwise, the connection could be timeout since compiling time increases with the growth of model size. +> RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/docs/programming_guide/en/master/distributed_training_ascend.html) , and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). For large models like InceptionV4, it's better to export an external environment variable `export HCCL_CONNECT_TIMEOUT=600` to extend hccl connection checking time from the default 120 seconds to 600 seconds. Otherwise, the connection could be timeout since compiling time increases with the growth of model size. > > This is processor cores binding operation regarding the `device_num` and total processor numbers. If you are not expect to do it, remove the operations `taskset` in `scripts/run_distribute_train.sh` > diff --git a/model_zoo/official/cv/densenet/README.md b/model_zoo/official/cv/densenet/README.md index d47ddb6c0f8..a724dbee656 100644 --- a/model_zoo/official/cv/densenet/README.md +++ b/model_zoo/official/cv/densenet/README.md @@ -75,7 +75,7 @@ The default configuration of the Dataset are as follows: ## Mixed Precision -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. diff --git a/model_zoo/official/cv/densenet/README_CN.md b/model_zoo/official/cv/densenet/README_CN.md index 846fb127d20..aa081d7e7be 100644 --- a/model_zoo/official/cv/densenet/README_CN.md +++ b/model_zoo/official/cv/densenet/README_CN.md @@ -80,7 +80,7 @@ DenseNet-100使用的数据集： Cifar-10 ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 # 环境要求 diff --git a/model_zoo/official/cv/dpn/README.md b/model_zoo/official/cv/dpn/README.md index 21a1350abba..e169974ce36 100644 --- a/model_zoo/official/cv/dpn/README.md +++ b/model_zoo/official/cv/dpn/README.md @@ -67,7 +67,7 @@ All the models in this repository are trained and validated on ImageNet-1K. The ## [Mixed Precision](#contents) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) diff --git a/model_zoo/official/cv/googlenet/README.md b/model_zoo/official/cv/googlenet/README.md index 98baf2ba4d3..b93cae220e0 100644 --- a/model_zoo/official/cv/googlenet/README.md +++ b/model_zoo/official/cv/googlenet/README.md @@ -66,7 +66,7 @@ Dataset used can refer to paper. ## Mixed Precision -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) diff --git a/model_zoo/official/cv/googlenet/README_CN.md b/model_zoo/official/cv/googlenet/README_CN.md index 7f47b6d8152..bdfff7d4ec4 100644 --- a/model_zoo/official/cv/googlenet/README_CN.md +++ b/model_zoo/official/cv/googlenet/README_CN.md @@ -20,7 +20,7 @@ - [分布式训练](#分布式训练) - [评估过程](#评估过程) - [评估](#评估) - - [导出过程](#导出过程) + - [导出过程](#导出过程) - [导出](#导出) - [推理过程](#推理过程) - [推理](#推理) @@ -73,7 +73,7 @@ GoogleNet由多个inception模块串联起来，可以更加深入。 降维的 ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 # 环境要求 @@ -84,7 +84,7 @@ GoogleNet由多个inception模块串联起来，可以更加深入。 降维的 - [MindSpore](https://www.mindspore.cn/install/en) - 如需查看详情，请参见如下资源： - [MindSpore教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) # 快速入门 @@ -549,7 +549,7 @@ python export.py --config_path [CONFIG_PATH] ### 推理 -如果您需要使用此训练模型在GPU、Ascend 910、Ascend 310等多个硬件平台上进行推理，可参考此[链接](https://www.mindspore.cn/docs/programming_guide/en/master/multi_platform_inference.html)。下面是操作步骤示例： +如果您需要使用此训练模型在GPU、Ascend 910、Ascend 310等多个硬件平台上进行推理，可参考此[链接](https://www.mindspore.cn/docs/programming_gui/zh-CN/master/multi_platform_inference.html)。下面是操作步骤示例： - Ascend处理器环境运行 diff --git a/model_zoo/official/cv/inceptionv3/README.md b/model_zoo/official/cv/inceptionv3/README.md index 4d11e255553..0d7ffcc12ac 100644 --- a/model_zoo/official/cv/inceptionv3/README.md +++ b/model_zoo/official/cv/inceptionv3/README.md @@ -52,7 +52,7 @@ Dataset used: [CIFAR-10](http://www.cs.toronto.edu/~kriz/cifar.html) ## [Mixed Precision(Ascend)](#contents) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. @@ -275,7 +275,7 @@ sh scripts/run_standalone_train.sh DEVICE_ID DATA_PATH sh scripts/run_standalone_train_cpu.sh DATA_PATH ``` -> Notes: RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/distributed_training_ascend.html), and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). For large models like InceptionV3, it's better to export an external environment variable `export HCCL_CONNECT_TIMEOUT=600` to extend hccl connection checking time from the default 120 seconds to 600 seconds. Otherwise, the connection could be timeout since compiling time increases with the growth of model size. +> Notes: RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/docs/programming_guide/en/master/distributed_training_ascend.html), and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). For large models like InceptionV3, it's better to export an external environment variable `export HCCL_CONNECT_TIMEOUT=600` to extend hccl connection checking time from the default 120 seconds to 600 seconds. Otherwise, the connection could be timeout since compiling time increases with the growth of model size. > > This is processor cores binding operation regarding the `device_num` and total processor numbers. If you are not expect to do it, remove the operations `taskset` in `scripts/run_distribute_train.sh` diff --git a/model_zoo/official/cv/inceptionv3/README_CN.md b/model_zoo/official/cv/inceptionv3/README_CN.md index 34d256cbecc..5bd1324a323 100644 --- a/model_zoo/official/cv/inceptionv3/README_CN.md +++ b/model_zoo/official/cv/inceptionv3/README_CN.md @@ -63,7 +63,7 @@ InceptionV3的总体网络架构如下： ## 混合精度（Ascend） -采用[混合精度](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 @@ -281,7 +281,7 @@ train.py和config.py中主要参数如下： sh scripts/run_standalone_train.sh DEVICE_ID DATA_PATH ``` -> 注：RANK_TABLE_FILE可参考[链接](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/distributed_training_ascend.html)。device_ip可以通过[链接](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools)获取 +> 注：RANK_TABLE_FILE可参考[链接](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/distributed_training_ascend.html)。device_ip可以通过[链接](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools)获取 > 这是关于device_num和处理器总数的处理器核绑定操作。如不需要，请删除scripts/run_distribute_train.sh中的taskset操作。 ### 启动 diff --git a/model_zoo/official/cv/inceptionv4/README.md b/model_zoo/official/cv/inceptionv4/README.md index ec7e9891a65..7e7c3550028 100644 --- a/model_zoo/official/cv/inceptionv4/README.md +++ b/model_zoo/official/cv/inceptionv4/README.md @@ -44,7 +44,7 @@ Dataset used can refer to paper. ## [Mixed Precision(Ascend)](#contents) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. @@ -253,7 +253,7 @@ sh scripts/run_standalone_train_ascend.sh DEVICE_ID DATA_DIR ``` > Notes: -> RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/distributed_training_ascend.html) , and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). For large models like InceptionV4, it's better to export an external environment variable `export HCCL_CONNECT_TIMEOUT=600` to extend hccl connection checking time from the default 120 seconds to 600 seconds. Otherwise, the connection could be timeout since compiling time increases with the growth of model size. +> RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/docs/programming_guide/en/master/distributed_training_ascend.html) , and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). For large models like InceptionV4, it's better to export an external environment variable `export HCCL_CONNECT_TIMEOUT=600` to extend hccl connection checking time from the default 120 seconds to 600 seconds. Otherwise, the connection could be timeout since compiling time increases with the growth of model size. > > This is processor cores binding operation regarding the `device_num` and total processor numbers. If you are not expect to do it, remove the operations `taskset` in `scripts/run_distribute_train.sh` diff --git a/model_zoo/official/cv/lenet_quant/Readme_CN.md b/model_zoo/official/cv/lenet_quant/Readme_CN.md index 6b4651b26d3..cc0852e25a0 100644 --- a/model_zoo/official/cv/lenet_quant/Readme_CN.md +++ b/model_zoo/official/cv/lenet_quant/Readme_CN.md @@ -63,10 +63,10 @@ LeNet非常简单，包含5层，由2个卷积层和3个全连接层组成。 - 硬件：Ascend - 使用Ascend搭建硬件环境 - 框架 - - [MindSpore](https://www.mindspore.cn/install/en) + - [MindSpore](https://www.mindspore.cn/install) - 如需查看详情，请参见如下资源： - - [MindSpore教程](https://www.mindspore.cn/tutorials/en/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) ## 快速入门 diff --git a/model_zoo/official/cv/maskrcnn/README.md b/model_zoo/official/cv/maskrcnn/README.md index dae335c3041..ea4c4cf2031 100644 --- a/model_zoo/official/cv/maskrcnn/README.md +++ b/model_zoo/official/cv/maskrcnn/README.md @@ -522,7 +522,7 @@ Usage: bash run_standalone_train.sh [PRETRAINED_MODEL] ## [Training Process](#contents) -- Set options in `config.py`, including loss_scale, learning rate and network hyperparameters. Click [here](https://www.mindspore.cn/tutorial/training/zh-CN/master/use/data_preparation.html) for more information about dataset. +- Set options in `config.py`, including loss_scale, learning rate and network hyperparameters. Click [here](https://www.mindspore.cn/docs/programming_guide/en/master/dataset_sample.html) for more information about dataset. ### [Training](#content) diff --git a/model_zoo/official/cv/maskrcnn/README_CN.md b/model_zoo/official/cv/maskrcnn/README_CN.md index cdf58e8e9ad..971e02ef0ae 100644 --- a/model_zoo/official/cv/maskrcnn/README_CN.md +++ b/model_zoo/official/cv/maskrcnn/README_CN.md @@ -517,7 +517,7 @@ bash run_eval.sh [VALIDATION_JSON_FILE] [CHECKPOINT_PATH] ## 训练过程 -- 在`config.py`中设置配置项，包括loss_scale、学习率和网络超参。单击[此处](https://www.mindspore.cn/tutorial/training/zh-CN/master/use/data_preparation.html)获取更多数据集相关信息. +- 在`config.py`中设置配置项，包括loss_scale、学习率和网络超参。单击[此处](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/dataset_sample.html)获取更多数据集相关信息. ### 训练 diff --git a/model_zoo/official/cv/maskrcnn_mobilenetv1/README.md b/model_zoo/official/cv/maskrcnn_mobilenetv1/README.md index 503901f0c9b..a1020a7f0a6 100644 --- a/model_zoo/official/cv/maskrcnn_mobilenetv1/README.md +++ b/model_zoo/official/cv/maskrcnn_mobilenetv1/README.md @@ -474,7 +474,7 @@ Usage: sh run_standalone_train.sh [PRETRAINED_MODEL] ## [Training Process](#contents) -- Set options in `config.py`, including loss_scale, learning rate and network hyperparameters. Click [here](https://www.mindspore.cn/tutorial/training/zh-CN/master/use/data_preparation.html) for more information about dataset. +- Set options in `config.py`, including loss_scale, learning rate and network hyperparameters. Click [here](https://www.mindspore.cn/docs/programming_guide/en/master/dataset_sample.html) for more information about dataset. ### [Training](#content) diff --git a/model_zoo/official/cv/mobilenetv1/README.md b/model_zoo/official/cv/mobilenetv1/README.md index 342c55b87b2..3fe662bb3cf 100644 --- a/model_zoo/official/cv/mobilenetv1/README.md +++ b/model_zoo/official/cv/mobilenetv1/README.md @@ -58,7 +58,7 @@ Dataset used: [ImageNet2012](http://www.image-net.org/) ### Mixed Precision(Ascend) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. ## Environment Requirements diff --git a/model_zoo/official/cv/mobilenetv2/README.md b/model_zoo/official/cv/mobilenetv2/README.md index de40982ad8c..61bd47f4de2 100644 --- a/model_zoo/official/cv/mobilenetv2/README.md +++ b/model_zoo/official/cv/mobilenetv2/README.md @@ -48,7 +48,7 @@ Dataset used: [imagenet](http://www.image-net.org/) ## [Mixed Precision(Ascend)](#contents) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) diff --git a/model_zoo/official/cv/mobilenetv2/README_CN.md b/model_zoo/official/cv/mobilenetv2/README_CN.md index 8ba4113143f..e52cd0ed106 100644 --- a/model_zoo/official/cv/mobilenetv2/README_CN.md +++ b/model_zoo/official/cv/mobilenetv2/README_CN.md @@ -54,7 +54,7 @@ MobileNetV2总体网络架构如下： ## 混合精度（Ascend） -采用[混合精度](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 # 环境要求 diff --git a/model_zoo/official/cv/mobilenetv2_quant/README_CN.md b/model_zoo/official/cv/mobilenetv2_quant/README_CN.md index 22382db63e6..94801c75dd3 100644 --- a/model_zoo/official/cv/mobilenetv2_quant/README_CN.md +++ b/model_zoo/official/cv/mobilenetv2_quant/README_CN.md @@ -59,7 +59,7 @@ MobileNetV2总体网络架构如下： ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html) 的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 diff --git a/model_zoo/official/cv/mobilenetv2_quant/Readme.md b/model_zoo/official/cv/mobilenetv2_quant/Readme.md index 8975d25c1f0..37c860f004f 100644 --- a/model_zoo/official/cv/mobilenetv2_quant/Readme.md +++ b/model_zoo/official/cv/mobilenetv2_quant/Readme.md @@ -46,7 +46,7 @@ Dataset used: [imagenet](http://www.image-net.org/) ## [Mixed Precision](#contents) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. ## [Learned Step Size Quantization](#contents) diff --git a/model_zoo/official/cv/nasnet/README_CN.md b/model_zoo/official/cv/nasnet/README_CN.md index 002471f769e..3451058bc53 100644 --- a/model_zoo/official/cv/nasnet/README_CN.md +++ b/model_zoo/official/cv/nasnet/README_CN.md @@ -49,8 +49,8 @@ NASNet总体网络架构如下： - [MindSpore](https://www.mindspore.cn/install) - 如需查看详情，请参见如下资源： - - [MindSpore教程](https://www.mindspore.cn/tutorials/en/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) # 脚本说明 diff --git a/model_zoo/official/cv/openpose/README.md b/model_zoo/official/cv/openpose/README.md index dc3f39f6fca..c928176b514 100644 --- a/model_zoo/official/cv/openpose/README.md +++ b/model_zoo/official/cv/openpose/README.md @@ -69,7 +69,7 @@ In the currently provided training script, the coco2017 data set is used as an e ## Mixed Precision -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) diff --git a/model_zoo/official/cv/resnet/README.md b/model_zoo/official/cv/resnet/README.md index b40408910a8..f83b39bda27 100644 --- a/model_zoo/official/cv/resnet/README.md +++ b/model_zoo/official/cv/resnet/README.md @@ -80,7 +80,7 @@ Dataset used: [ImageNet2012](http://www.image-net.org/) ## Mixed Precision -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data types, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data types, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) @@ -373,7 +373,7 @@ bash run_eval_gpu_resnet_benchmark.sh [DATASET_PATH] [CKPT_PATH] [BATCH_SIZE](op For distributed training, a hostfile configuration needs to be created in advance. -Please follow the instructions in the link [GPU-Multi-Host](https://www.mindspore.cn/tutorial/training/zh-CN/r1.0/advanced_use/distributed_training_gpu.html). +Please follow the instructions in the link [GPU-Multi-Host](https://www.mindspore.cn/docs/programming_guide/en/master/distributed_training_gpu.html). #### Running parameter server mode training diff --git a/model_zoo/official/cv/resnet/README_CN.md b/model_zoo/official/cv/resnet/README_CN.md index 872b61b52d0..f26e7ef7967 100755 --- a/model_zoo/official/cv/resnet/README_CN.md +++ b/model_zoo/official/cv/resnet/README_CN.md @@ -83,7 +83,7 @@ ResNet的总体网络架构如下： ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 # 环境要求 diff --git a/model_zoo/official/cv/resnet50_quant/README.md b/model_zoo/official/cv/resnet50_quant/README.md index 940d231fe99..3121326008d 100644 --- a/model_zoo/official/cv/resnet50_quant/README.md +++ b/model_zoo/official/cv/resnet50_quant/README.md @@ -53,7 +53,7 @@ Dataset used: [ImageNet2012](http://www.image-net.org/) ## [Mixed Precision](#contents) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) diff --git a/model_zoo/official/cv/resnet50_quant/README_CN.md b/model_zoo/official/cv/resnet50_quant/README_CN.md index 6cad2cf9fe6..90b5a560091 100644 --- a/model_zoo/official/cv/resnet50_quant/README_CN.md +++ b/model_zoo/official/cv/resnet50_quant/README_CN.md @@ -58,7 +58,7 @@ ResNet-50总体网络架构如下： ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 # 环境要求 @@ -71,9 +71,9 @@ ResNet-50总体网络架构如下： - 如需查看详情，请参见如下资源： - - [MindSpore教程](https://www.mindspore.cn/tutorials/en/master/index.html) + - [MindSpore教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) ## 脚本说明 diff --git a/model_zoo/official/cv/resnext/README.md b/model_zoo/official/cv/resnext/README.md index f7cbeb78b92..c3661de4220 100644 --- a/model_zoo/official/cv/resnext/README.md +++ b/model_zoo/official/cv/resnext/README.md @@ -47,7 +47,7 @@ Dataset used: [imagenet](http://www.image-net.org/) ## [Mixed Precision](#contents) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. diff --git a/model_zoo/official/cv/resnext/README_CN.md b/model_zoo/official/cv/resnext/README_CN.md index 6becbf04842..635dcfdd709 100644 --- a/model_zoo/official/cv/resnext/README_CN.md +++ b/model_zoo/official/cv/resnext/README_CN.md @@ -54,7 +54,7 @@ ResNeXt整体网络架构如下： ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 diff --git a/model_zoo/official/cv/retinanet/README_CN.md b/model_zoo/official/cv/retinanet/README_CN.md index 833ed0300ae..73a91d0756a 100644 --- a/model_zoo/official/cv/retinanet/README_CN.md +++ b/model_zoo/official/cv/retinanet/README_CN.md @@ -67,10 +67,10 @@ MSCOCO2017 - 硬件（Ascend） - 使用Ascend处理器准备硬件环境。 - 架构 - - [MindSpore](https://www.mindspore.cn/install/en) + - [MindSpore](https://www.mindspore.cn/install) - 想要获取更多信息，请检查以下资源： - - [MindSpore 教程](https://www.mindspore.cn/tutorials/en/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore 教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) ## [脚本说明](#content) @@ -190,7 +190,7 @@ sh scripts/run_single_train.sh DEVICE_ID MINDRECORD_DIR PRE_TRAINED(optional) PR > 注意: - RANK_TABLE_FILE相关参考资料见[链接](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/distributed_training_ascend.html), 获取device_ip方法详见[链接](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). + RANK_TABLE_FILE相关参考资料见[链接](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/distributed_training_ascend.html), 获取device_ip方法详见[链接](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). #### 运行 diff --git a/model_zoo/official/cv/simple_pose/README.md b/model_zoo/official/cv/simple_pose/README.md index 08cd16ed436..419208e55e3 100644 --- a/model_zoo/official/cv/simple_pose/README.md +++ b/model_zoo/official/cv/simple_pose/README.md @@ -57,7 +57,7 @@ Dataset used: COCO2017 ## [Mixed Precision](#contents) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) diff --git a/model_zoo/official/cv/squeezenet/README.md b/model_zoo/official/cv/squeezenet/README.md index 6814356d735..b09159fb52b 100644 --- a/model_zoo/official/cv/squeezenet/README.md +++ b/model_zoo/official/cv/squeezenet/README.md @@ -62,7 +62,7 @@ Dataset used: [ImageNet2012](http://www.image-net.org/) ## Mixed Precision -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) diff --git a/model_zoo/official/cv/ssd/README.md b/model_zoo/official/cv/ssd/README.md index 633d2fb58c0..421c2bfea97 100644 --- a/model_zoo/official/cv/ssd/README.md +++ b/model_zoo/official/cv/ssd/README.md @@ -302,7 +302,7 @@ Then you can run everything just like on ascend. ### [Training Process](#contents) -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), `voc_root`(voc dataset) or `image_dir` and `anno_path`(own dataset). **Note if mindrecord_dir isn't empty, it will use mindrecord_dir instead of raw images.** +To train the model, run `train.py`. If the `mindrecord_dir` is empty, it will generate [mindrecord](https://www.mindspore.cn/docs/programming_guide/en/master/convert_dataset.html) files by `coco_root`(coco dataset), `voc_root`(voc dataset) or `image_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 diff --git a/model_zoo/official/cv/ssd/README_CN.md b/model_zoo/official/cv/ssd/README_CN.md index b7e5015af49..039d496292b 100644 --- a/model_zoo/official/cv/ssd/README_CN.md +++ b/model_zoo/official/cv/ssd/README_CN.md @@ -246,7 +246,7 @@ sh run_eval_gpu.sh [DATASET] [CHECKPOINT_PATH] [DEVICE_ID] [CONFIG_PATH] ## 训练过程 -运行`train.py`训练模型。如果`mindrecord_dir`为空，则会通过`coco_root`（coco数据集）或`image_dir`和`anno_path`（自己的数据集）生成[MindRecord](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/convert_dataset.html)文件。**注意，如果mindrecord_dir不为空，将使用mindrecord_dir代替原始图像。** +运行`train.py`训练模型。如果`mindrecord_dir`为空，则会通过`coco_root`（coco数据集）或`image_dir`和`anno_path`（自己的数据集）生成[MindRecord](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/convert_dataset.html)文件。**注意，如果mindrecord_dir不为空，将使用mindrecord_dir代替原始图像。** ### Ascend上训练 diff --git a/model_zoo/official/cv/tinydarknet/README_CN.md b/model_zoo/official/cv/tinydarknet/README_CN.md index 48f0cbd66f1..3943f32c1f9 100644 --- a/model_zoo/official/cv/tinydarknet/README_CN.md +++ b/model_zoo/official/cv/tinydarknet/README_CN.md @@ -64,17 +64,17 @@ Tiny-DarkNet是Joseph Chet Redmon等人提出的一个16层的针对于经典的 - + # [环境要求](#目录) - 硬件（Ascend/CPU） - 请准备具有Ascend/CPU处理器的硬件环境. - 框架 - - [MindSpore](https://www.mindspore.cn/install/en) + - [MindSpore](https://www.mindspore.cn/install) - 更多的信息请访问以下链接： - - [MindSpore Tutorials](https://www.mindspore.cn/tutorials/en/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore Tutorials](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) # [快速入门](#目录) diff --git a/model_zoo/official/cv/unet3d/README.md b/model_zoo/official/cv/unet3d/README.md index a7ad1e07a03..a5cdac62ce7 100644 --- a/model_zoo/official/cv/unet3d/README.md +++ b/model_zoo/official/cv/unet3d/README.md @@ -281,7 +281,7 @@ After training, you'll get some checkpoint files under the `train_parallel_fp[32 #### Distributed training on Ascend > Notes: -> RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/distributed_training_ascend.html) , and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). For large models like InceptionV4, it's better to export an external environment variable `export HCCL_CONNECT_TIMEOUT=600` to extend hccl connection checking time from the default 120 seconds to 600 seconds. Otherwise, the connection could be timeout since compiling time increases with the growth of model size. +> RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/docs/programming_guide/en/master/distributed_training_ascend.html) , and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). For large models like InceptionV4, it's better to export an external environment variable `export HCCL_CONNECT_TIMEOUT=600` to extend hccl connection checking time from the default 120 seconds to 600 seconds. Otherwise, the connection could be timeout since compiling time increases with the growth of model size. > ```shell diff --git a/model_zoo/official/cv/vgg16/README.md b/model_zoo/official/cv/vgg16/README.md index 2780f1956af..87ccec30949 100644 --- a/model_zoo/official/cv/vgg16/README.md +++ b/model_zoo/official/cv/vgg16/README.md @@ -84,7 +84,7 @@ Note that you can run the scripts based on the dataset mentioned in original pap ### Mixed Precision -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. @@ -445,7 +445,7 @@ train_parallel1/log:epcoh: 2 step: 97, loss is 1.7133579 ... ``` -> About rank_table.json, you can refer to the [distributed training tutorial](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/distributed_training_tutorials.html). +> About rank_table.json, you can refer to the [distributed training tutorial](https://www.mindspore.cn/docs/programming_guide/en/master/distributed_training.html). > **Attention** This will bind the processor cores according to the `device_num` and total processor numbers. If you don't expect to run pretraining with binding processor cores, remove the operations about `taskset` in `scripts/run_distribute_train.sh` ##### Run vgg16 on GPU diff --git a/model_zoo/official/cv/vgg16/README_CN.md b/model_zoo/official/cv/vgg16/README_CN.md index 2096c19bcd3..294ceee5c2f 100644 --- a/model_zoo/official/cv/vgg16/README_CN.md +++ b/model_zoo/official/cv/vgg16/README_CN.md @@ -87,7 +87,7 @@ VGG 16网络主要由几个基本模块（包括卷积层和池化层）和三 ### 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 @@ -449,7 +449,7 @@ train_parallel1/log:epcoh: 2 step: 97, loss is 1.7133579 ... ``` -> 关于rank_table.json，可以参考[分布式并行训练](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/distributed_training_tutorials.html)。 +> 关于rank_table.json，可以参考[分布式并行训练](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/distributed_training.html)。 > **注意** 将根据`device_num`和处理器总数绑定处理器核。如果您不希望预训练中绑定处理器内核，请在`scripts/run_distribute_train.sh`脚本中移除`taskset`相关操作。 ##### GPU处理器环境运行VGG16 diff --git a/model_zoo/official/cv/warpctc/README.md b/model_zoo/official/cv/warpctc/README.md index 3fb4d6da38d..5d8b392a39d 100644 --- a/model_zoo/official/cv/warpctc/README.md +++ b/model_zoo/official/cv/warpctc/README.md @@ -244,7 +244,7 @@ save_checkpoint_path: "./checkpoint" # path to save checkpoint ### [Training Process](#contents) -- Set options in `default_config.yaml`, including learning rate and other network hyperparameters. Click [MindSpore dataset preparation tutorial](https://www.mindspore.cn/tutorial/training/zh-CN/master/use/data_preparation.html) for more information about dataset. +- Set options in `default_config.yaml`, including learning rate and other network hyperparameters. Click [MindSpore dataset preparation tutorial](https://www.mindspore.cn/docs/programming_guide/en/master/dataset_sample.html) for more information about dataset. #### [Training](#contents) diff --git a/model_zoo/official/cv/warpctc/README_CN.md b/model_zoo/official/cv/warpctc/README_CN.md index 7f2226ac913..37727c6ea3f 100644 --- a/model_zoo/official/cv/warpctc/README_CN.md +++ b/model_zoo/official/cv/warpctc/README_CN.md @@ -248,7 +248,7 @@ save_checkpoint_path: "./checkpoints" # 检查点保存路径，相对于t ## 训练过程 -- 在`default_config.yaml`中设置选项，包括学习率和网络超参数。单击[MindSpore加载数据集教程](https://www.mindspore.cn/tutorial/training/zh-CN/master/use/data_preparation.html)，了解更多信息。 +- 在`default_config.yaml`中设置选项，包括学习率和网络超参数。单击[MindSpore加载数据集教程](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/dataset_sample.html)，了解更多信息。 ### 训练 diff --git a/model_zoo/official/cv/xception/README.md b/model_zoo/official/cv/xception/README.md index 4beb4f44b93..cb6b751dca7 100644 --- a/model_zoo/official/cv/xception/README.md +++ b/model_zoo/official/cv/xception/README.md @@ -54,7 +54,7 @@ Dataset used can refer to paper. ## [Mixed Precision](#contents) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. @@ -193,7 +193,7 @@ sh run_eval_gpu.sh DEVICE_ID DATASET_PATH CHECKPOINT_PATH sh run_infer_310.sh MINDIR_PATH DATA_PATH LABEL_FILE DEVICE_ID ``` -> Notes: RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/distributed_training_ascend.html), and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). +> Notes: RANK_TABLE_FILE can refer to [Link](https://www.mindspore.cn/docs/programming_guide/en/master/distributed_training_ascend.html), and the device_ip can be got as [Link](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). ### Launch diff --git a/model_zoo/official/cv/yolov3_resnet18/README.md b/model_zoo/official/cv/yolov3_resnet18/README.md index f6df768e5e1..5d420840b32 100644 --- a/model_zoo/official/cv/yolov3_resnet18/README.md +++ b/model_zoo/official/cv/yolov3_resnet18/README.md @@ -263,7 +263,7 @@ After installing MindSpore via the official website, you can start training and ### Training on Ascend -To train the model, run `train.py` with the dataset `image_dir`, `anno_path` and `mindrecord_dir`. If the `mindrecord_dir` is empty, it wil generate [mindrecord](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/convert_dataset.html) file by `image_dir` and `anno_path`(the absolute image path is joined by the `image_dir` and the relative path in `anno_path`). **Note if `mindrecord_dir` isn't empty, it will use `mindrecord_dir` rather than `image_dir` and `anno_path`.** +To train the model, run `train.py` with the dataset `image_dir`, `anno_path` and `mindrecord_dir`. If the `mindrecord_dir` is empty, it wil generate [mindrecord](https://www.mindspore.cn/docs/programming_guide/en/master/convert_dataset.html) file by `image_dir` and `anno_path`(the absolute image path is joined by the `image_dir` and the relative path in `anno_path`). **Note if `mindrecord_dir` isn't empty, it will use `mindrecord_dir` rather than `image_dir` and `anno_path`.** - Stand alone mode @@ -304,7 +304,7 @@ Note the results is two-classification(person and face) used our own annotations ### Evaluation on Ascend -To eval, run `eval.py` with the dataset `image_dir`, `anno_path`(eval txt), `mindrecord_dir` and `ckpt_path`. `ckpt_path` is the path of [checkpoint](https://www.mindspore.cn/tutorial/training/en/master/use/save_model.html) file. +To eval, run `eval.py` with the dataset `image_dir`, `anno_path`(eval txt), `mindrecord_dir` and `ckpt_path`. `ckpt_path` is the path of [checkpoint](https://www.mindspore.cn/docs/programming_guide/en/master/save_model.html) file. ```bash sh run_eval.sh 0 yolo.ckpt ./Mindrecord_eval ./dataset ./dataset/eval.txt diff --git a/model_zoo/official/cv/yolov3_resnet18/README_CN.md b/model_zoo/official/cv/yolov3_resnet18/README_CN.md index 3264e305e7a..8dd749eb65c 100644 --- a/model_zoo/official/cv/yolov3_resnet18/README_CN.md +++ b/model_zoo/official/cv/yolov3_resnet18/README_CN.md @@ -266,7 +266,7 @@ YOLOv3整体网络架构如下： ### Ascend上训练 -训练模型运行`train.py`，使用数据集`image_dir`、`anno_path`和`mindrecord_dir`。如果`mindrecord_dir`为空，则通过`image_dir`和`anno_path`（图像绝对路径由`image_dir`和`anno_path`中的相对路径连接）生成[MindRecord](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/convert_dataset.html)文件。**注意，如果`mindrecord_dir`不为空，将使用`mindrecord_dir`而不是`image_dir`和`anno_path`。** +训练模型运行`train.py`，使用数据集`image_dir`、`anno_path`和`mindrecord_dir`。如果`mindrecord_dir`为空，则通过`image_dir`和`anno_path`（图像绝对路径由`image_dir`和`anno_path`中的相对路径连接）生成[MindRecord](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/convert_dataset.html)文件。**注意，如果`mindrecord_dir`不为空，将使用`mindrecord_dir`而不是`image_dir`和`anno_path`。** - 单机模式 @@ -307,7 +307,7 @@ YOLOv3整体网络架构如下： ### Ascend评估 -运行`eval.py`，数据集为`image_dir`、`anno_path`(评估TXT)、`mindrecord_dir`和`ckpt_path`。`ckpt_path`是[检查点](https://www.mindspore.cn/tutorial/training/zh-CN/master/use/save_model.html)文件的路径。 +运行`eval.py`，数据集为`image_dir`、`anno_path`(评估TXT)、`mindrecord_dir`和`ckpt_path`。`ckpt_path`是[检查点](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/save_model.html)文件的路径。 ```shell script sh run_eval.sh 0 yolo.ckpt ./Mindrecord_eval ./dataset ./dataset/eval.txt diff --git a/model_zoo/official/nlp/bert/README.md b/model_zoo/official/nlp/bert/README.md index ba5bde7772a..4cf07cc7cdf 100644 --- a/model_zoo/official/nlp/bert/README.md +++ b/model_zoo/official/nlp/bert/README.md @@ -193,7 +193,7 @@ For distributed training among multiple machines, training command should be exe Please follow the instructions in the link below to create an hccl.json file in need: [https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). -For dataset, if you want to set the format and parameters, a schema configuration file with JSON format needs to be created, please refer to [tfrecord](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/dataset_loading.html#tfrecord) format. +For dataset, if you want to set the format and parameters, a schema configuration file with JSON format needs to be created, please refer to [tfrecord](https://www.mindspore.cn/docs/programming_guide/en/master/dataset_loading.html#tfrecord) format. ```text For pretraining, schema file contains ["input_ids", "input_mask", "segment_ids", "next_sentence_labels", "masked_lm_positions", "masked_lm_ids", "masked_lm_weights"]. diff --git a/model_zoo/official/nlp/cpm/README.md b/model_zoo/official/nlp/cpm/README.md index 1b6d087a9d7..d958bf510f9 100644 --- a/model_zoo/official/nlp/cpm/README.md +++ b/model_zoo/official/nlp/cpm/README.md @@ -52,8 +52,8 @@ CPM is implemented by GPT, which includes multi-layer decoder module. - Framework - [MindSpore](https://gitee.com/mindspore/mindspore) - For more information, please check the resources below: - - [MindSpore Tutorials](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) + - [MindSpore Tutorials](https://www.mindspore.cn/tutorials/en/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) # Quick Start @@ -309,7 +309,7 @@ After processing, the mindrecord file of training and reasoning is generated in ### Finetune Training Process -- Set options in `src/config.py`, including loss_scale, learning rate and network hyperparameters. Click [here](https://www.mindspore.cn/tutorial/training/zh-CN/master/use/data_preparation.html) for more information about dataset. +- Set options in `src/config.py`, including loss_scale, learning rate and network hyperparameters. Click [here](https://www.mindspore.cn/docs/programming_guide/en/master/dataset_sample.html) for more information about dataset. - Run `run_distribute_train_ascend_single_machine.sh` for distributed and single machine training of CPM model. diff --git a/model_zoo/official/nlp/cpm/README_CN.md b/model_zoo/official/nlp/cpm/README_CN.md index 6de41f5a0d1..739fc1aec5f 100644 --- a/model_zoo/official/nlp/cpm/README_CN.md +++ b/model_zoo/official/nlp/cpm/README_CN.md @@ -309,7 +309,7 @@ Parameters for dataset and network (Training/Evaluation): ### Finetune训练过程 -- 在`src/config.py`中设置，包括模型并行、batchsize、学习率和网络超参数。点击[这里](https://www.mindspore.cn/tutorial/training/zh-CN/master/use/data_preparation.html)查看更多数据集信息。 +- 在`src/config.py`中设置，包括模型并行、batchsize、学习率和网络超参数。点击[这里](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/dataset_sample.html)查看更多数据集信息。 - 运行`run_distribute_train_ascend_single_machine.sh`，进行CPM模型的单机8卡分布式训练。 diff --git a/model_zoo/official/nlp/emotect/README_CN.md b/model_zoo/official/nlp/emotect/README_CN.md index 345d1d87c20..ba9464b293a 100755 --- a/model_zoo/official/nlp/emotect/README_CN.md +++ b/model_zoo/official/nlp/emotect/README_CN.md @@ -56,10 +56,10 @@ label text_a - 硬件（Ascend/GPU） - 使用Ascend或GPU处理器来搭建硬件环境。 - 框架 - - [MindSpore](https://www.mindspore.cn/install/en) + - [MindSpore](https://www.mindspore.cn/install) - 如需查看详情，请参见如下资源： - [MindSpore教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) # 快速入门 diff --git a/model_zoo/official/nlp/pangu_alpha/README.md b/model_zoo/official/nlp/pangu_alpha/README.md index 317f39ab52e..9e2222445ee 100644 --- a/model_zoo/official/nlp/pangu_alpha/README.md +++ b/model_zoo/official/nlp/pangu_alpha/README.md @@ -27,7 +27,7 @@ with our parallel setting. We summarized the training tricks as followings: 2. Pipeline Model Parallelism 3. Optimizer Model Parallelism -The above features can be found [here](https://www.mindspore.cn/doc/programming_guide/en/r1.2/auto_parallel.html). +The above features can be found [here](https://www.mindspore.cn/docs/programming_guide/en/master/auto_parallel.html). More amazing features are still under developing. The technical report and checkpoint file can be found [here](https://git.openi.org.cn/PCL-Platform.Intelligence/PanGu-AIpha). @@ -133,7 +133,7 @@ bash scripts/run_distribute_training.sh DATASET RANK_TABLE RANK_SIZE TYPE MODE The above command involves some `args` described below: - DATASET: The path to the mindrecord files's parent directory . For example: `/home/work/mindrecord/`. -- RANK_TABLE: The details of the rank table can be found [here](https://www.mindspore.cn/tutorial/training/zh-CN/r1.2/advanced_use/distributed_training_ascend.html). It's a json file describes the `device id`, `service ip` and `rank`. +- RANK_TABLE: The details of the rank table can be found [here](https://www.mindspore.cn/docs/programming_guide/en/master/distributed_training_ascend.html). It's a json file describes the `device id`, `service ip` and `rank`. - RANK_SIZE: The device number. This can be your total device numbers. For example, 8, 16, 32 ... - TYPE: The param init type. The parameters will be initialized with float32. Or you can replace it with `fp16`. This will save a little memory used on the device. - MODE: The configure mode. This mode will set the `hidden size` and `layers` to make the parameter number near 2.6 billions. The other mode can be `13B` (`hidden size` 5120 and `layers` 40, which needs at least 16 cards to train.) and `200B`. @@ -159,7 +159,7 @@ bash scripts/run_distributed_train_gpu.sh RANK_SIZE HOSTFILE DATASET MOD ``` - RANK_SIZE: The device number. This can be your total device numbers. For example, 8, 16, 32 ... -- HOSTFILE: It's a text file describes the host ip and its devices. Please see our [tutorial](https://www.mindspore.cn/tutorial/training/zh-CN/r1.2/advanced_use/distributed_training_gpu.html) or [OpenMPI](https://www.open-mpi.org/) for more details. +- HOSTFILE: It's a text file describes the host ip and its devices. Please see our [tutorial](https://www.mindspore.cn/docs/programming_guide/en/master/distributed_training_gpu.html) or [OpenMPI](https://www.open-mpi.org/) for more details. - DATASET: The path to the mindrecord files's parent directory . For example: `/home/work/mindrecord/`. - MODE: Can be `2.6B`, `13B` and `200B`. diff --git a/model_zoo/official/nlp/transformer/README.md b/model_zoo/official/nlp/transformer/README.md index 6834d9e2538..7fb17d2586a 100644 --- a/model_zoo/official/nlp/transformer/README.md +++ b/model_zoo/official/nlp/transformer/README.md @@ -309,7 +309,7 @@ Parameters for learning rate: ## [Training Process](#contents) -- Set options in `default_config.yaml`, including loss_scale, learning rate and network hyperparameters. Click [here](https://www.mindspore.cn/tutorial/training/zh-CN/master/use/data_preparation.html) for more information about dataset. +- Set options in `default_config.yaml`, including loss_scale, learning rate and network hyperparameters. Click [here](https://www.mindspore.cn/docs/programming_guide/en/master/dataset_sample.html) for more information about dataset. - Run `run_standalone_train.sh` for non-distributed training of Transformer model. diff --git a/model_zoo/official/nlp/transformer/README_CN.md b/model_zoo/official/nlp/transformer/README_CN.md index 910f17d9431..aecf3cf4dc4 100644 --- a/model_zoo/official/nlp/transformer/README_CN.md +++ b/model_zoo/official/nlp/transformer/README_CN.md @@ -316,7 +316,7 @@ Parameters for learning rate: ### 训练过程 -- 在`default_config.yaml`中设置选项，包括loss_scale、学习率和网络超参数。点击[这里](https://www.mindspore.cn/tutorial/training/zh-CN/master/use/data_preparation.html)查看更多数据集信息。 +- 在`default_config.yaml`中设置选项，包括loss_scale、学习率和网络超参数。点击[这里](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/dataset_sample.html)查看更多数据集信息。 - 运行`run_standalone_train.sh`，进行Transformer模型的非分布式训练。 diff --git a/model_zoo/official/recommend/ncf/README.md b/model_zoo/official/recommend/ncf/README.md index a4f2f44d2ac..69c2c0153b8 100644 --- a/model_zoo/official/recommend/ncf/README.md +++ b/model_zoo/official/recommend/ncf/README.md @@ -73,7 +73,7 @@ In both datasets, the timestamp is represented in seconds since midnight Coordin ## Mixed Precision -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) @@ -310,7 +310,7 @@ Inference result is saved in current path, you can find result like this in acc. If you need to use the trained model to perform inference on multiple hardware platforms, such as Ascend 910 or Ascend 310, you can refer to this [Link](https://www.mindspore.cn/docs/programming_guide/en/master/multi_platform_inference.html). Following the steps below, this is a simple example: - + ```python # Load unseen dataset for inference diff --git a/model_zoo/research/audio/fcn-4/README.md b/model_zoo/research/audio/fcn-4/README.md index 8b0e3bc2c22..cf2407d491f 100644 --- a/model_zoo/research/audio/fcn-4/README.md +++ b/model_zoo/research/audio/fcn-4/README.md @@ -36,7 +36,7 @@ FCN-4 is a convolutional neural network architecture, its name FCN-4 comes from ### Mixed Precision -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. ## [Environment Requirements](#contents) diff --git a/model_zoo/research/cv/AttGAN/README_CN.md b/model_zoo/research/cv/AttGAN/README_CN.md index 3a7be73b220..c90cd2d1790 100644 --- a/model_zoo/research/cv/AttGAN/README_CN.md +++ b/model_zoo/research/cv/AttGAN/README_CN.md @@ -53,10 +53,10 @@ CelebFaces Attributes Dataset (CelebA) 是一个大规模的人脸属性数据 - 硬件（Ascend） - 使用Ascend来搭建硬件环境。 - 框架 - - [MindSpore](https://www.mindspore.cn/install/en) + - [MindSpore](https://www.mindspore.cn/install) - 如需查看详情，请参见如下资源： - [MindSpore教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) # 快速入门 diff --git a/model_zoo/research/cv/GENet_Res50/README_CN.md b/model_zoo/research/cv/GENet_Res50/README_CN.md index d5811a292ca..ee2ddfd8fa0 100644 --- a/model_zoo/research/cv/GENet_Res50/README_CN.md +++ b/model_zoo/research/cv/GENet_Res50/README_CN.md @@ -60,7 +60,7 @@ Imagenet 2017和Imagenet 2012 数据集一致 ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 # 环境要求 @@ -73,9 +73,9 @@ Imagenet 2017和Imagenet 2012 数据集一致 - 如需查看详情，请参见如下资源： - - [MindSpore教程](https://www.mindspore.cn/tutorials/en/master/index.html) + - [MindSpore教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) ## 脚本说明 diff --git a/model_zoo/research/cv/ManiDP/Readme.md b/model_zoo/research/cv/ManiDP/Readme.md index 3dbdc60d749..dc2ffdde050 100755 --- a/model_zoo/research/cv/ManiDP/Readme.md +++ b/model_zoo/research/cv/ManiDP/Readme.md @@ -37,7 +37,7 @@ Dataset used: [CIFAR-10](https://www.cs.toronto.edu/~kriz/cifar.html) ## [Mixed Precision(Ascend)](#contents) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) @@ -47,7 +47,7 @@ For FP16 operators, if the input data type is FP32, the backend of MindSpore wil - Framework - [MindSpore](https://www.mindspore.cn/install/en) - For more information, please check the resources below： - - [MindSpore tutorials](https://www.mindspore.cn/tutorial/zh-CN/master/index.html) + - [MindSpore tutorials](https://www.mindspore.cn/tutorial/en/master/index.html) - [MindSpore API](https://www.mindspore.cn/docs/api/en/master/index.html) # [Script description](#contents) diff --git a/model_zoo/research/cv/SE-Net/README.md b/model_zoo/research/cv/SE-Net/README.md index 1bf311cad4b..f23d6f37512 100644 --- a/model_zoo/research/cv/SE-Net/README.md +++ b/model_zoo/research/cv/SE-Net/README.md @@ -53,7 +53,7 @@ Dataset used: [ImageNet2012](http://www.image-net.org/) ## Mixed Precision -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data types, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data types, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) diff --git a/model_zoo/research/cv/arcface/README.md b/model_zoo/research/cv/arcface/README.md index aa578c7b4a7..2bf302531a0 100644 --- a/model_zoo/research/cv/arcface/README.md +++ b/model_zoo/research/cv/arcface/README.md @@ -213,7 +213,7 @@ epoch time: 1104929.793 ms, per step time: 97.162 ms ### 推理 -如果您需要使用已训练模型在GPU、Ascend 910、Ascend 310等多个硬件平台上进行推理，可参考[此处](https://www.mindspore.cn/tutorial/inference/zh-CN/r1.2/index.html)。 +如果您需要使用已训练模型在GPU、Ascend 910、Ascend 310等多个硬件平台上进行推理，可参考[此处](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/multi_platform_inference.html)。 ### 迁移学习 diff --git a/model_zoo/research/cv/deeplabv3plus/README_CN.md b/model_zoo/research/cv/deeplabv3plus/README_CN.md index 0af05a61663..886a0abe1d4 100644 --- a/model_zoo/research/cv/deeplabv3plus/README_CN.md +++ b/model_zoo/research/cv/deeplabv3plus/README_CN.md @@ -79,7 +79,7 @@ Pascal VOC数据集和语义边界数据集（Semantic Boundaries Dataset，SBD ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 # 环境要求 diff --git a/model_zoo/research/cv/glore_res200/README_CN.md b/model_zoo/research/cv/glore_res200/README_CN.md index 42673d684b0..b555915ae9e 100644 --- a/model_zoo/research/cv/glore_res200/README_CN.md +++ b/model_zoo/research/cv/glore_res200/README_CN.md @@ -65,7 +65,7 @@ ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_gui/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 # 环境要求 diff --git a/model_zoo/research/cv/glore_res50/README.md b/model_zoo/research/cv/glore_res50/README.md index 357fc8094de..b4af3299325 100644 --- a/model_zoo/research/cv/glore_res50/README.md +++ b/model_zoo/research/cv/glore_res50/README.md @@ -61,14 +61,14 @@ glore_res的总体网络架构如下： ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 # 环境要求 - 硬件(Ascend) - 框架 - - [MindSpore](https://www.mindspore.cn/install/en) + - [MindSpore](https://www.mindspore.cn/install) - 如需查看详情，请参见如下资源： - [MindSpore教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) diff --git a/model_zoo/research/cv/hardnet/README_CN.md b/model_zoo/research/cv/hardnet/README_CN.md index fdf1eed005f..b9eb10bdd63 100644 --- a/model_zoo/research/cv/hardnet/README_CN.md +++ b/model_zoo/research/cv/hardnet/README_CN.md @@ -60,7 +60,7 @@ HarDNet指的是Harmonic DenseNet: A low memory traffic network，其突出的 ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 # 环境要求 @@ -68,10 +68,10 @@ HarDNet指的是Harmonic DenseNet: A low memory traffic network，其突出的 - 硬件（Ascend/GPU） - 使用Ascend或GPU处理器来搭建硬件环境。 - 框架 - - [MindSpore](https://www.mindspore.cn/install/en) + - [MindSpore](https://www.mindspore.cn/install) - 如需查看详情，请参见如下资源： - [MindSpore教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) # 快速入门 @@ -419,7 +419,7 @@ bash run_infer_310.sh [MINDIR_PATH] [DATASET_PATH] [DEVICE_ID] ### 推理 -如果您需要使用此训练模型在Ascend 910上进行推理，可参考此[链接](https://www.mindspore.cn/docs/programming_guide/en/master/multi_platform_inference.html)。下面是操作步骤示例： +如果您需要使用此训练模型在Ascend 910上进行推理，可参考此[链接](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/multi_platform_inference.html)。下面是操作步骤示例： - Ascend处理器环境运行 @@ -456,7 +456,7 @@ bash run_infer_310.sh [MINDIR_PATH] [DATASET_PATH] [DEVICE_ID] print("==============Acc: {} ==============".format(acc)) ``` -如果您需要使用此训练模型在GPU上进行推理，可参考此[链接](https://www.mindspore.cn/docs/programming_guide/en/master/multi_platform_inference.html)。下面是操作步骤示例： +如果您需要使用此训练模型在GPU上进行推理，可参考此[链接](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/multi_platform_inference.html)。下面是操作步骤示例： - GPU处理器环境运行 diff --git a/model_zoo/research/cv/ntsnet/README.md b/model_zoo/research/cv/ntsnet/README.md index 608695f5056..4b3f15d16dd 100644 --- a/model_zoo/research/cv/ntsnet/README.md +++ b/model_zoo/research/cv/ntsnet/README.md @@ -126,7 +126,7 @@ Usage: bash run_standalone_train.sh [DATA_URL] [TRAIN_URL] ## [Training Process](#contents) -- Set options in `config.py`, including learning rate, output filename and network hyperparameters. Click [here](https://www.mindspore.cn/tutorial/training/zh-CN/master/use/data_preparation.html) for more information about dataset. +- Set options in `config.py`, including learning rate, output filename and network hyperparameters. Click [here](https://www.mindspore.cn/docs/programming_guide/en/master/dataset_sample.html) for more information about dataset. ### [Training](#content) diff --git a/model_zoo/research/cv/renas/Readme.md b/model_zoo/research/cv/renas/Readme.md index 0ddf1cd398e..6bb56593ca7 100755 --- a/model_zoo/research/cv/renas/Readme.md +++ b/model_zoo/research/cv/renas/Readme.md @@ -36,7 +36,7 @@ An effective and efficient architecture performance evaluation scheme is essenti ## [Mixed Precision(Ascend)](#contents) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) @@ -46,7 +46,7 @@ For FP16 operators, if the input data type is FP32, the backend of MindSpore wil - Framework - [MindSpore](https://www.mindspore.cn/install/en) - For more information, please check the resources below： - - [MindSpore tutorials](https://www.mindspore.cn/tutorial/zh-CN/master/index.html) + - [MindSpore tutorials](https://www.mindspore.cn/tutorial/en/master/index.html) - [MindSpore API](https://www.mindspore.cn/docs/api/en/master/index.html) # [Script description](#contents) diff --git a/model_zoo/research/cv/retinanet_resnet101/README_CN.md b/model_zoo/research/cv/retinanet_resnet101/README_CN.md index 03b99040e0b..236a26884cb 100644 --- a/model_zoo/research/cv/retinanet_resnet101/README_CN.md +++ b/model_zoo/research/cv/retinanet_resnet101/README_CN.md @@ -60,10 +60,10 @@ MSCOCO2017 - 硬件（Ascend） - 使用Ascend处理器准备硬件环境。 - 架构 - - [MindSpore](https://www.mindspore.cn/install/en) + - [MindSpore](https://www.mindspore.cn/install) - 想要获取更多信息，请检查以下资源： - - [MindSpore 教程](https://www.mindspore.cn/tutorials/en/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore 教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) ## [脚本说明](#content) @@ -178,7 +178,7 @@ sh run_distribute_train.sh DEVICE_ID EPOCH_SIZE LR DATASET PRE_TRAINED(optional) > 注意: - RANK_TABLE_FILE相关参考资料见[链接](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/distributed_training_ascend.html), 获取device_ip方法详见[链接](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). + RANK_TABLE_FILE相关参考资料见[链接](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/distributed_training_ascend.html), 获取device_ip方法详见[链接](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). #### 运行 diff --git a/model_zoo/research/cv/retinanet_resnet152/README_CN.md b/model_zoo/research/cv/retinanet_resnet152/README_CN.md index 9629f6a89a7..69d33e4461f 100644 --- a/model_zoo/research/cv/retinanet_resnet152/README_CN.md +++ b/model_zoo/research/cv/retinanet_resnet152/README_CN.md @@ -60,10 +60,10 @@ MSCOCO2017 - 硬件（Ascend） - 使用Ascend处理器准备硬件环境。 - 架构 - - [MindSpore](https://www.mindspore.cn/install/en) + - [MindSpore](https://www.mindspore.cn/install) - 想要获取更多信息，请检查以下资源： - - [MindSpore 教程](https://www.mindspore.cn/tutorials/en/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore 教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) ## [脚本说明](#content) @@ -178,7 +178,7 @@ sh run_distribute_train.sh DEVICE_ID EPOCH_SIZE LR DATASET PRE_TRAINED(optional) > 注意: - RANK_TABLE_FILE相关参考资料见[链接](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/distributed_training_ascend.html), 获取device_ip方法详见[链接](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). + RANK_TABLE_FILE相关参考资料见[链接](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/distributed_training_ascend.html), 获取device_ip方法详见[链接](https://gitee.com/mindspore/mindspore/tree/master/model_zoo/utils/hccl_tools). #### 运行 diff --git a/model_zoo/research/cv/squeezenet/README.md b/model_zoo/research/cv/squeezenet/README.md index c227069efac..1ff80dea40a 100644 --- a/model_zoo/research/cv/squeezenet/README.md +++ b/model_zoo/research/cv/squeezenet/README.md @@ -57,7 +57,7 @@ Dataset used: [ImageNet2012](http://www.image-net.org/) ## Mixed Precision -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) diff --git a/model_zoo/research/cv/ssd_ghostnet/README.md b/model_zoo/research/cv/ssd_ghostnet/README.md index 47d7fcd0c27..fa0b9f39e6e 100644 --- a/model_zoo/research/cv/ssd_ghostnet/README.md +++ b/model_zoo/research/cv/ssd_ghostnet/README.md @@ -200,7 +200,7 @@ If you want to run in modelarts, please check the official documentation of [mod ### 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.** +To train the model, run `train.py`. If the `mindrecord_dir` is empty, it will generate [mindrecord](https://www.mindspore.cn/docs/programming_guide/en/master/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.** - Distribute mode diff --git a/model_zoo/research/cv/ssd_mobilenetV2/README.md b/model_zoo/research/cv/ssd_mobilenetV2/README.md index 3bb07a51b92..1dfd99ed1f2 100644 --- a/model_zoo/research/cv/ssd_mobilenetV2/README.md +++ b/model_zoo/research/cv/ssd_mobilenetV2/README.md @@ -186,7 +186,7 @@ sh scripts/run_eval.sh [DATASET] [CHECKPOINT_PATH] [DEVICE_ID] ### [Training Process](#contents) -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), `voc_root`(voc dataset) or `image_dir` and `anno_path`(own dataset). **Note if mindrecord_dir isn't empty, it will use mindrecord_dir instead of raw images.** +To train the model, run `train.py`. If the `mindrecord_dir` is empty, it will generate [mindrecord](https://www.mindspore.cn/docs/programming_guide/en/master/convert_dataset.html) files by `coco_root`(coco dataset), `voc_root`(voc dataset) or `image_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 diff --git a/model_zoo/research/cv/ssd_mobilenetV2_FPNlite/README.md b/model_zoo/research/cv/ssd_mobilenetV2_FPNlite/README.md index 623fba720de..fe946b51b63 100644 --- a/model_zoo/research/cv/ssd_mobilenetV2_FPNlite/README.md +++ b/model_zoo/research/cv/ssd_mobilenetV2_FPNlite/README.md @@ -190,7 +190,7 @@ sh scripts/run_eval.sh [DATASET] [CHECKPOINT_PATH] [DEVICE_ID] ### [Training Process](#contents) -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), `voc_root`(voc dataset) or `image_dir` and `anno_path`(own dataset). **Note if mindrecord_dir isn't empty, it will use mindrecord_dir instead of raw images.** +To train the model, run `train.py`. If the `mindrecord_dir` is empty, it will generate [mindrecord](https://www.mindspore.cn/docs/programming_guide/en/master/convert_dataset.html) files by `coco_root`(coco dataset), `voc_root`(voc dataset) or `image_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 diff --git a/model_zoo/research/cv/ssd_resnet50/README.md b/model_zoo/research/cv/ssd_resnet50/README.md index def4d8812bb..febc0023d6f 100644 --- a/model_zoo/research/cv/ssd_resnet50/README.md +++ b/model_zoo/research/cv/ssd_resnet50/README.md @@ -204,7 +204,7 @@ Then you can run everything just like on ascend. ### [Training Process](#contents) -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), `voc_root`(voc dataset) or `image_dir` and `anno_path`(own dataset). **Note if mindrecord_dir isn't empty, it will use mindrecord_dir instead of raw images.** +To train the model, run `train.py`. If the `mindrecord_dir` is empty, it will generate [mindrecord](https://www.mindspore.cn/docs/programming_guide/en/master/convert_dataset.html) files by `coco_root`(coco dataset), `voc_root`(voc dataset) or `image_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 diff --git a/model_zoo/research/cv/ssd_resnet50/README_CN.md b/model_zoo/research/cv/ssd_resnet50/README_CN.md index d4ae9f07fcf..663d8746a65 100644 --- a/model_zoo/research/cv/ssd_resnet50/README_CN.md +++ b/model_zoo/research/cv/ssd_resnet50/README_CN.md @@ -163,7 +163,7 @@ sh run_eval.sh [DATASET] [CHECKPOINT_PATH] [DEVICE_ID] ## 训练过程 -运行`train.py`训练模型。如果`mindrecord_dir`为空，则会通过`coco_root`（coco数据集）或`image_dir`和`anno_path`（自己的数据集）生成[MindRecord](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/convert_dataset.html)文件。**注意，如果mindrecord_dir不为空，将使用mindrecord_dir代替原始图像。** +运行`train.py`训练模型。如果`mindrecord_dir`为空，则会通过`coco_root`（coco数据集）或`image_dir`和`anno_path`（自己的数据集）生成[MindRecord](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/convert_dataset.html)文件。**注意，如果mindrecord_dir不为空，将使用mindrecord_dir代替原始图像。** ### Ascend上训练 diff --git a/model_zoo/research/cv/vgg19/README_CN.md b/model_zoo/research/cv/vgg19/README_CN.md index 8561a83871c..edf7c811e21 100644 --- a/model_zoo/research/cv/vgg19/README_CN.md +++ b/model_zoo/research/cv/vgg19/README_CN.md @@ -68,7 +68,7 @@ VGG 19网络主要由几个基本模块（包括卷积层和池化层）和三 ## 混合精度 -采用[混合精度](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 diff --git a/model_zoo/research/cv/wgan/README_CN.md b/model_zoo/research/cv/wgan/README_CN.md index b10d1b365d1..dc01521b8f8 100644 --- a/model_zoo/research/cv/wgan/README_CN.md +++ b/model_zoo/research/cv/wgan/README_CN.md @@ -55,10 +55,10 @@ WGAN网络包含两部分，生成器网络和判别器网络。判别器网络 - 硬件（Ascend） - 使用Ascend来搭建硬件环境。 - 框架 - - [MindSpore](https://www.mindspore.cn/install/en) + - [MindSpore](https://www.mindspore.cn/install) - 如需查看详情，请参见如下资源： - [MindSpore教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) # 快速入门 diff --git a/model_zoo/research/hpc/pinns/README.md b/model_zoo/research/hpc/pinns/README.md index c564895caca..c49dbed00b2 100644 --- a/model_zoo/research/hpc/pinns/README.md +++ b/model_zoo/research/hpc/pinns/README.md @@ -72,7 +72,7 @@ Dataset used：[cylinder nektar wake](https://github.com/maziarraissi/PINNs/tree ## [Mixed Precision](#Contents) -The [mixed precision](https://www.mindspore.cn/tutorial/training/en/master/advanced_use/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. +The [mixed precision](https://www.mindspore.cn/docs/programming_guide/en/master/enable_mixed_precision.html) training method accelerates the deep learning neural network training process by using both the single-precision and half-precision data formats, and maintains the network precision achieved by the single-precision training at the same time. Mixed precision training can accelerate the computation process, reduce memory usage, and enable a larger model or batch size to be trained on specific hardware. For FP16 operators, if the input data type is FP32, the backend of MindSpore will automatically handle it with reduced precision. Users could check the reduced-precision operators by enabling INFO log and then searching ‘reduce precision’. # [Environment Requirements](#contents) diff --git a/model_zoo/research/hpc/pinns/README_CN.md b/model_zoo/research/hpc/pinns/README_CN.md index 6ec189848cb..e6850496115 100644 --- a/model_zoo/research/hpc/pinns/README_CN.md +++ b/model_zoo/research/hpc/pinns/README_CN.md @@ -70,7 +70,7 @@ Navier-Stokes方程是流体力学中描述粘性牛顿流体的方程。针对N ## [混合精度](#目录) -采用[混合精度](https://www.mindspore.cn/tutorial/training/zh-CN/master/advanced_use/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 +采用[混合精度](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/enable_mixed_precision.html)的训练方法使用支持单精度和半精度数据来提高深度学习神经网络的训练速度，同时保持单精度训练所能达到的网络精度。混合精度训练提高计算速度、减少内存使用的同时，支持在特定硬件上训练更大的模型或实现更大批次的训练。 以FP16算子为例，如果输入数据类型为FP32，MindSpore后台会自动降低精度来处理数据。用户可打开INFO日志，搜索“reduce precision”查看精度降低的算子。 # [环境要求](#目录) @@ -78,10 +78,10 @@ Navier-Stokes方程是流体力学中描述粘性牛顿流体的方程。针对N - 硬件（GPU） - 使用GPU处理器来搭建硬件环境。 - 框架 - - [MindSpore](https://www.mindspore.cn/install/en) + - [MindSpore](https://www.mindspore.cn/install) - 如需查看详情，请参见如下资源： - [MindSpore教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html) - - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html) + - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html) # [快速入门](#目录) diff --git a/model_zoo/research/hpc/sponge/README.md b/model_zoo/research/hpc/sponge/README.md index 83c82d21b54..5c0ab5cee64 100644 --- a/model_zoo/research/hpc/sponge/README.md +++ b/model_zoo/research/hpc/sponge/README.md @@ -26,7 +26,7 @@ There are three inputs in the example, the property file `NVT_290_10ns.in`, the The topology file and coordinates file can be generated by `tleap` in `AmberTools` ([link]()). For more details, please refer to: -- [SPONGE Tutorial](https://gitee.com/mindspore/docs/blob/master/tutorials/training/source_zh_cn/advanced_use/hpc_sponge.md) +- [SPONGE Tutorial](https://gitee.com/mindspore/docs/blob/master/docs/mindspore/programming_guide/source_en/hpc_sponge.md) ## Environment Requirements