From 16572f279098d2447feb182cf71c64073caa2580 Mon Sep 17 00:00:00 2001 From: LDOUBLEV Date: Wed, 10 Jun 2020 20:44:06 +0800 Subject: [PATCH] add doc of decommpressing pretrain weight file --- doc/doc_ch/detection.md | 12 ++++++++++++ doc/doc_en/detection_en.md | 12 ++++++++++++ 2 files changed, 24 insertions(+) diff --git a/doc/doc_ch/detection.md b/doc/doc_ch/detection.md index 849755bf..fbff7075 100644 --- a/doc/doc_ch/detection.md +++ b/doc/doc_ch/detection.md @@ -43,6 +43,18 @@ cd PaddleOCR/ wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV3_large_x0_5_pretrained.tar # 下载ResNet50的预训练模型 wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_ssld_pretrained.tar + +# 解压预训练模型文件，以MobileNetV3为例 +tar xf ./pretrain_models/MobileNetV3_large_x0_5_pretrained.tar ./pretrain_models/ + +# 注：正确解压backbone预训练权重文件后，文件夹下包含众多以网络层命名的权重文件，格式如下： +./pretrain_models/MobileNetV3_large_x0_5_pretrained/ + └─ conv_last_bn_mean + └─ conv_last_bn_offset + └─ conv_last_bn_scale + └─ conv_last_bn_variance + └─ ...... + ``` **启动训练** diff --git a/doc/doc_en/detection_en.md b/doc/doc_en/detection_en.md index eb500879..619f2c01 100644 --- a/doc/doc_en/detection_en.md +++ b/doc/doc_en/detection_en.md @@ -42,6 +42,18 @@ cd PaddleOCR/ wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV3_large_x0_5_pretrained.tar # Download the pre-trained model of ResNet50 wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_ssld_pretrained.tar + +# decompressing the pre-training model file, take MobileNetV3 as an example +tar xf ./pretrain_models/MobileNetV3_large_x0_5_pretrained.tar ./pretrain_models/ + +# Note: After decompressing the backbone pre-training weight file correctly, the file list in the folder is as follows: +./pretrain_models/MobileNetV3_large_x0_5_pretrained/ + └─ conv_last_bn_mean + └─ conv_last_bn_offset + └─ conv_last_bn_scale + └─ conv_last_bn_variance + └─ ...... + ``` **Start training** -- GitLab