From 96dae875f32fb88dd52e8dc5892f55dfff4ecfdd Mon Sep 17 00:00:00 2001 From: Jay Wong Date: Thu, 30 Apr 2020 17:01:29 -0400 Subject: [PATCH] Fix typo --- README.md | 6 +++--- tiny-vgg/README.md | 8 ++++---- 2 files changed, 7 insertions(+), 7 deletions(-) diff --git a/README.md b/README.md index 837b80fe..860aec2d 100644 --- a/README.md +++ b/README.md @@ -17,7 +17,7 @@ Clone or download this repository: ``` git clone git@github.com:poloclub/cnn-explainer.git -# Or use degit, if you don't want to download commit histories +# Use degit if you don't want to download commit histories degit poloclub/cnn-explainer ``` @@ -27,13 +27,13 @@ Install the dependencies: npm install ``` -Than run CNN Explainer: +Then run CNN Explainer: ``` npm run dev ``` -Navigate to [localhost:5000](https://localhost:5000). You should CNN Explainer running in your broswer :) +Navigate to [localhost:5000](https://localhost:5000). You should see CNN Explainer running in your broswer :) To see how we trained the CNN, visit the directory [`./tiny-vgg/`](tiny-vgg). diff --git a/tiny-vgg/README.md b/tiny-vgg/README.md index 0242ed25..e1dff7c4 100644 --- a/tiny-vgg/README.md +++ b/tiny-vgg/README.md @@ -2,7 +2,7 @@ This directory includes code and data to train a Tiny VGG model (inspired by the demo CNN in [Stanford CS231n class](http://cs231n.stanford.edu)) -on 10 every-day classes from the [Tiny ImageNet](https://tiny-imagenet.herokuapp.com). +on 10 everyday classes from the [Tiny ImageNet](https://tiny-imagenet.herokuapp.com). ## Installation @@ -55,7 +55,7 @@ First, you want to unzip `data.zip`. The file structure would be something like: To install all dependencies, run the following code ``` -conda env create --file environment.yml +conda env create --file environment.yaml ``` ## Training @@ -66,14 +66,14 @@ To train Tiny VGG on these 10 classes, run the following code python tiny-vgg.py ``` -After training, you will get two trained models in Keras format `trained_tiny_vgg.h5` +After training, you will get two saved models in Keras format: `trained_tiny_vgg.h5` and `trained_vgg_best.h5`. The first file is the final model after training, and `trained_vgg_best.h5` is the model having the best validation performance. You can use either one for CNN Explainer. ## Convert Model Format -Before loading the model using *tensorflow.js*, you want to convert the model format +Before loading the model using *tensorflow.js*, you want to convert the model file from Keras `h5` format to [tensorflow.js format](https://www.tensorflow.org/js/tutorials/conversion/import_keras). ```