final update for report

Author: isaaccorley

README.md

@@ -1,2 +1,45 @@
 # pytorch-modulation-recognition
 PyTorch Implementation of various Modulation Recognition Networks benchmarked on the RadioML2016 Dataset
+
+## Install Requirements
+
+```bash
+# Base requirements
+pip3 install -r requirements.txt
+
+# Required to use train.py
+pip3 install poutyne
+
+# Required to run the demo web app
+pip3 install streamlit
+
+```
+
+## Train
+
+```bash
+# Train VTCNN2
+python train.py --model vtcnn --epochs 25 --batch_size 512 --split 0.8
+
+# Train MRResNet
+python train.py --model mrresnet --epochs 25 --batch_size 512 --split 0.8
+
+```
+
+### Results
+
+After training Tensorboard logs will be located in the logs directory and the models in the models directory. Run the below command to start Tensorboard and point your browser to localhost:6006.
+
+```bash
+tensorboard --logdir=logs
+
+```
+
+### Demo
+
+To run the web app for visualizing the In-Phase (I) and Quadrature (Q) channels of signals at various modulations and signal-to-noise ratios, use the following command.
+
+```bash
+streamlit run app.py
+
+```

demo/app.py renamed to app.py

@@ -4,7 +4,6 @@
 import matplotlib.pyplot as plt
 import streamlit as st
 
-sys.path.append("../")
 import torch_modulation_recognition as tmr
 
 
@@ -18,7 +17,7 @@
 
 @st.cache(allow_output_mutation=True)
 def load_data():
-    data = tmr.RadioML2016()
+    data = tmr.data.RadioML2016()
     return data
 
 def get_params(data):
@@ -58,7 +57,7 @@ def main(data):
     if params["BUTTON_PRESSED"]:
         fig, ax = plt.subplots(nrows=2, ncols=1, sharey=True, sharex=True)
         plt.xlabel("time (s)")
-        fig.suptitle("({}) Modulation: {}, Signal-to-Noise Ratio: {}".format(params["IDX"], params["MOD"], params["SNR"]))
+        fig.suptitle("({}) Modulation: {}, Signal-to-Noise Ratio: {} (In-Phase (I) and Quadrature (Q) Channels)".format(params["IDX"], params["MOD"], params["SNR"]))
         ax[0].plot(params["SIGNALS"][(params["MOD"], params["SNR"])][params["IDX"], 0, :])
         ax[1].plot(params["SIGNALS"][(params["MOD"], params["SNR"])][params["IDX"], 1, :])
         st.pyplot()

requirements.txt

@@ -1,5 +1,3 @@
 numpy
 torch
-poutyne
-tqdm
-streamlit
+tqdm