add images

Author: ttduongtran

README.MD

@@ -1,4 +1,5 @@
 # GRU
+Implemention for Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling. [Read paper](https://arxiv.org/pdf/1412.3555.pdf)
 
 Library Logo <--- **FIXME**
 
@@ -11,9 +12,12 @@ Compare and evaluate recurrent neural networks (RNN) with different types of rec
 - Long short-term memory (LSTM)
 - Tanh
 
-Slide about your project (if it's available) <--- **FIXME**
+Slide about our project [vietnamese](https://docs.google.com/document/d/1GHZIDAyfCL2O-PVPVyQJdcQ8FxdkPGA4-AQd3cCYoMQ/edit?usp=sharing)
 
-Architecture Image  <--- **FIXME**
+Architecture
+<p align="center">
+    <img src='./assets/architecture.png' width="90%" class="center">
+</p>
 
 Advisors:
 - Github: [bangoc123](https://github.com/bangoc123)
@@ -187,22 +191,22 @@ Epoch 10/10
 
 ```
 
-### 4. Comments about these results
+### 4. Results
 
-- Base on the results of Training with 3 kinds of model, we can figure out that there are differences about the Accuaracy between them
+- Base on the results of Training with 3 kinds of model, we can figure out that there are differences about the Accuracy between them
 	+ Both GRU and LSTM have high efficiency with the highest val_accuracy is 0.8936 and 0.8795 , respectively.    
 	+ Tanh is the opposite, the val_accuracy is really low with just around 0.54 
 - The results clearly demonstrated the superiority of the gated units; both the LSTM unit and GRU, over the traditional tanh unit.
 - However, we could not make concrete conclusion on which of the two gating units was better.
 
-## VI. Running Test
+<!-- ## VI. Running Test
 
 When you want to modify the model, you need to run the test to make sure your change does not affect the whole system.
 
 In the `./folder-name` **(FIXME)** folder please run:
 
 ```bash
 pytest
-```
+``` -->
 
 

assets/architecture.png

@@ -1,10 +1,10 @@
 import os
 from argparse import ArgumentParser
 import tensorflow as tf
-from model.gru_rnn import GRU_RNN
 from data import Dataset
 import pandas as pd
 import numpy as np
+os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
 
 if __name__ == "__main__":
     parser = ArgumentParser()

assets/gru.png

@@ -123,6 +123,6 @@
     model.save(f"{args.model_folder}/{args.model}.h5py")
 
     # Do Prediction
-    print('==============Evaluate=============')
-    model.evaluate(val_ds, batch_size=128)
+    # print('==============Evaluate=============')
+    # model.evaluate(val_ds, batch_size=128)