aadityaknv / mahe_workshop_image_classifier Goto Github PK

READ ME

**Download and Install Anaconda** *(https://www.anaconda.com/distribution/)*

**1. Once it is installed, open the Anaconda prompt and install the following things**

```
conda create --name tensorflow python=3.6
```
**Windows:**
```
activate tensorflow
```
**Mac:**
```
source activate tensorflow
```


```
python retrain.py --image_dir=data_dir \
--saved_model_dir=saved_model_dir \
--bottleneck_dir=bottleneck_dir \
--how_many_training_steps=2000 \
--output_labels=output/output_labels.txt \
--output_graph=output/retrain.pb
```

```
conda install scipy
pip install --upgrade sklearn
pip install --upgrade pandas
pip install --upgrade pandas-datareader
pip install --upgrade matplotlib
pip install --upgrade pillow
pip install --upgrade tqdm
pip install --upgrade requests
pip install --upgrade h5py
pip install --upgrade pyyaml
pip install --upgrade tensorflow_hub
pip install --upgrade bayesian-optimization
pip install --upgrade tensorflow==1.12.0
pip install --upgrade keras==2.2.4
```


**Download Link for Flower Dataset :** http://download.tensorflow.org/example_images/flower_photos.tgz

**Retrain Command :** python -m retrain --bottleneck_dir=tf_files/bottlenecks --model_dir=tf_files/models/"${ARCHITECTURE}" --summaries_dir=tf_files/training_summaries/"${ARCHITECTURE}" --output_graph=tf_files/retrained_graph.pb --output_labels=tf_files/retrained_labels.txt --architecture="${ARCHITECTURE}" --image_dir=tf_files/flower_photos