Spaces:
Runtime error
Runtime error
Commit
•
cc50161
1
Parent(s):
ba8a301
hosting RNN Playgroung on hugging face
Browse files- .gitattributes +3 -0
- app.py +184 -0
- button_style.css +3 -0
- info.jpg +3 -0
- predict.py +152 -0
- requirements.txt +6 -0
- wait.gif +3 -0
.gitattributes
CHANGED
@@ -32,3 +32,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
32 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
33 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
34 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
32 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
33 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
34 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
35 |
+
*.jpg filter=lfs diff=lfs merge=lfs -text
|
36 |
+
*.gif filter=lfs diff=lfs merge=lfs -text
|
37 |
+
*.css filter=lfs diff=lfs merge=lfs -text
|
app.py
ADDED
@@ -0,0 +1,184 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
import numpy as np
|
3 |
+
import matplotlib.pyplot as plt
|
4 |
+
import seaborn as sns
|
5 |
+
import requests
|
6 |
+
import json
|
7 |
+
|
8 |
+
from predict import predict_series
|
9 |
+
|
10 |
+
st.set_page_config(page_title='RNN Playground')
|
11 |
+
#st.set_option('deprecation.showPyplotGlobalUse', False)
|
12 |
+
|
13 |
+
pages = {'Intro': 0, 'Implementation details': 1, 'The model': 2}
|
14 |
+
choice = pages[st.sidebar.radio("Select the chapter: ", tuple(pages.keys()))]
|
15 |
+
|
16 |
+
|
17 |
+
if choice == 0:
|
18 |
+
st.title("Recurrent Neural Networks playground")
|
19 |
+
st.subheader("The purpose")
|
20 |
+
st.write("""\n
|
21 |
+
The goal of this app is to allow the user to experiment easily with Recurrent Neural Networks. Thanks to that, the app helps to understand: \n
|
22 |
+
- When to use recurrent neural networks
|
23 |
+
- Which scenarios are more straightforward for the RNN to predict and which are more difficult
|
24 |
+
- How the noise interrupts the predictions
|
25 |
+
- Difference between LSTM and GRU nodes
|
26 |
+
- Understanding that the increasing number of nodes doesn't always lead to better performance
|
27 |
+
""")
|
28 |
+
|
29 |
+
st.subheader("Typical use case")
|
30 |
+
st.write("""
|
31 |
+
1. Create a synthetic dataset, with a wide range of many choices of parameters
|
32 |
+
2. Create a Recurrent Neural Networks model by selecting a number of nodes in particular layers of the model
|
33 |
+
3. Automatically train the RNN model and make predictions
|
34 |
+
4. Compare the predicted values with the actual values
|
35 |
+
\n""")
|
36 |
+
st.subheader('The architecture of the model')
|
37 |
+
st.image('info.jpg', use_column_width=True, caption='Hover the cursor over image to see the enlarge button')
|
38 |
+
st.write(""" \n Use the radio buttons on the left to navigate between chapters \n \n \n \n""")
|
39 |
+
|
40 |
+
elif choice == 1:
|
41 |
+
|
42 |
+
st.title(""" \n Implementation details""")
|
43 |
+
st.subheader("Front-end")
|
44 |
+
st.write("""\n
|
45 |
+
The front-end part was made with the use of [the Streamlit library](https://www.streamlit.io/).
|
46 |
+
The parameters from the sidebar are used to create a dataset. The dataset is visualised using the Seaborn
|
47 |
+
library and finally sent (with the parameters specifying the number of neurons in particular layers) to
|
48 |
+
the back-end part through REST API. \n
|
49 |
+
The front-end is served on Azure as a Web App. """)
|
50 |
+
|
51 |
+
st.subheader("Back-end")
|
52 |
+
st.subheader("[Since the playground is stored on HuggingFace now - the backend is a module of the frontend]")
|
53 |
+
st.write("""\n
|
54 |
+
|
55 |
+
The backed-part is responsible for:
|
56 |
+
- Retrieving the dataset from the front-end through REST API
|
57 |
+
- Creating an RNN model using parameters passed from the user
|
58 |
+
- Training the model
|
59 |
+
- Predicting the values and returning them to the front-end \n
|
60 |
+
|
61 |
+
The most crucial requirements are:
|
62 |
+
- The neural networks setup has to be able to accurately predict the further shape of a curve for
|
63 |
+
the widest range of parameters selected by the user.
|
64 |
+
- Time execution of the back-end part must be short.
|
65 |
+
- Which means balancing over tradeoff between the time needed for the response and the accuracy of the results
|
66 |
+
- Cost efficiency
|
67 |
+
- Since the app is desired to be on-line all the time, the serverless approach has been taken.
|
68 |
+
That's why the back-end is served on Azure as a serverless Function App.
|
69 |
+
|
70 |
+
""")
|
71 |
+
|
72 |
+
else:
|
73 |
+
|
74 |
+
gran = 0.25
|
75 |
+
test_len = 8
|
76 |
+
st.sidebar.header('User Input Parameters')
|
77 |
+
|
78 |
+
|
79 |
+
def user_input_features():
|
80 |
+
predefined_sets = {'length': [30, ], 'period': [1.34, ], 'amplitude': [0.64, ], 'growth': [0.04, ],
|
81 |
+
'amplitude_growth': [0.03, ], 'r1_nodes': [20, ], 'r2_nodes': [20, ], 'fc1_nodes': [34, ]}
|
82 |
+
|
83 |
+
data, nn = {}, {}
|
84 |
+
st.sidebar.header('Dataset:')
|
85 |
+
data['length'] = st.sidebar.slider('Training data length', 20, 50, 28)
|
86 |
+
data['period'] = st.sidebar.slider('Period of the wave', 0.75, 2.0, 1.0)
|
87 |
+
data['growth'] = st.sidebar.slider('Values growth', -0.25, 0.25, 0.0)
|
88 |
+
data['amplitude'] = st.sidebar.slider('Amplitude', 0.25, 1.75, 1.0)
|
89 |
+
data['amplitude_growth'] = st.sidebar.slider('Amplitude growth', -0.01, 0.1, 0.0)
|
90 |
+
data['noise'] = st.sidebar.slider('Noise', 0.0, 1.0, 0.0)
|
91 |
+
st.sidebar.header('Model setup')
|
92 |
+
nn['use_lstm'] = st.sidebar.radio('Select the type of Recurrent Neuron to use', ['LSTM', 'GRU']) == 'LSTM'
|
93 |
+
nn['r1_nodes'] = st.sidebar.slider('Number of nodes in the first RNN layer', 1, 30, 13)
|
94 |
+
nn['r2_nodes'] = st.sidebar.slider('Number of nodes in the second RNN layer', 0, 30, 0)
|
95 |
+
nn['fc1_nodes'] = st.sidebar.slider('Number of nodes in the fully connected RNN layer', 0, 40, 10)
|
96 |
+
nn['steps'] = len(np.arange(0, test_len, gran))
|
97 |
+
|
98 |
+
#if st.sidebar.button('Load one of the pretested configurations'):
|
99 |
+
#i = st.sidebar.selectbox('Select:', [-1, 0])
|
100 |
+
#i = int(np.random.rand(len(predefined_sets['length']))) # Selecting one pretested configuration
|
101 |
+
#data.update({k: predefined_sets[k][i] for k in set(data) & set(predefined_sets)})
|
102 |
+
#nn.update({k: predefined_sets[k][i] for k in set(nn) & set(predefined_sets)})"""
|
103 |
+
|
104 |
+
|
105 |
+
return data, nn
|
106 |
+
|
107 |
+
|
108 |
+
params, setup = user_input_features()
|
109 |
+
|
110 |
+
st.subheader("Instructions:")
|
111 |
+
st.write("""
|
112 |
+
1. Modify the dataset by using the sliders in the Dataset group on the left on the screen.
|
113 |
+
2. Select the number of nodes in the model by using the sliders in the RNN setup group.
|
114 |
+
3. Press the "Train and Predict" button to Train and Predict the model - note: many operations performing under the hood - please be patient.
|
115 |
+
4. The predicted values will be shown at the bottom of the page.
|
116 |
+
5. If you are not satisfied with the results - modify the model and try again!
|
117 |
+
6. Have fun!
|
118 |
+
\n""")
|
119 |
+
|
120 |
+
st.subheader("Generated data:")
|
121 |
+
X = np.arange(0, params['length'], gran)
|
122 |
+
X_pred = np.arange(params['length'], params['length'] + test_len, gran)
|
123 |
+
|
124 |
+
|
125 |
+
def generate_wave(x_set):
|
126 |
+
return np.sin(x_set / params['period']) * (1 + params['amplitude_growth'] * x_set) * params[
|
127 |
+
'amplitude'] + x_set * params['growth'] + params['noise']*np.random.randn(len(x_set))
|
128 |
+
|
129 |
+
|
130 |
+
Y = generate_wave(X)
|
131 |
+
Y_pred = generate_wave(X_pred)
|
132 |
+
|
133 |
+
X_pred, Y_pred = np.append(X[-1], X_pred), np.append(Y[-1], Y_pred)
|
134 |
+
|
135 |
+
c1, c2, c3 = '#1e4a76', '#7dc0f7', '#ff7c0a' # colors
|
136 |
+
# sns.scatterplot(x=X, y=Y, color=c1)
|
137 |
+
# st.pyplot()
|
138 |
+
sns.lineplot(x=X, y=Y, color=c1)
|
139 |
+
sns.lineplot(x=X_pred, y=Y_pred, color=c2, linestyle=':')
|
140 |
+
plt.ylim(min(-2, min(Y), min(Y_pred)), max(2, max(Y), max(Y_pred)))
|
141 |
+
plt.legend(['Train data', 'Test data'], loc=3)
|
142 |
+
plt.xlabel('Sample number')
|
143 |
+
plt.ylabel('Sample value')
|
144 |
+
st.pyplot()
|
145 |
+
st.write("The plot presents generated train and test data. Use the sliders on the left to modify the curve.")
|
146 |
+
|
147 |
+
|
148 |
+
def local_css(file_name):
|
149 |
+
with open(file_name) as f:
|
150 |
+
st.markdown(f'<style>{f.read()}</style>', unsafe_allow_html=True)
|
151 |
+
local_css("button_style.css")
|
152 |
+
|
153 |
+
st.subheader('Predicted data:')
|
154 |
+
reminder = st.text('Press the train and predict button on the sidebar once you are ready with the selections.')
|
155 |
+
|
156 |
+
|
157 |
+
if st.sidebar.button('Train and Predict'):
|
158 |
+
setup['values'] = list(Y)
|
159 |
+
reminder.empty()
|
160 |
+
|
161 |
+
waiters = list()
|
162 |
+
waiters.append(st.text('Please wait till the train and predict process is finished.'))
|
163 |
+
waiters.append(st.image('wait.gif'))
|
164 |
+
waiters.append(st.text("""The process should take around 20-60 seconds."""))
|
165 |
+
|
166 |
+
# myUrl = 'http://localhost:7071/api/predict'
|
167 |
+
myUrl = 'https://rnn-background.azurewebsites.net/api/predict?code=a/X0yioXXY4CFVd9UFTw4MiyStNJ2qh3oae7FdFN7VBFMFhqe/qK7Q=='
|
168 |
+
# request = json.dumps(setup)
|
169 |
+
result = predict_series(data, steps=66, r1_nodes=14, r2_nodes=14, fc1_nodes=20)
|
170 |
+
# _ = [waiter.empty() for waiter in waiters]
|
171 |
+
|
172 |
+
result = np.array(range(5))
|
173 |
+
|
174 |
+
sns.lineplot(x=X_pred, y=Y_pred, color=c2, linestyle=':')
|
175 |
+
sns.lineplot(x=X, y=Y, color=c1)
|
176 |
+
sns.lineplot(np.append(X[-1], np.arange(0, test_len, gran) + max(X) + gran), np.append(Y[-1], result['result']), color=c3)
|
177 |
+
plt.legend(['Train data', 'Test data', 'Predicted data'], loc=3)
|
178 |
+
plt.xlabel('Sample number')
|
179 |
+
plt.ylabel('Sample value')
|
180 |
+
st.pyplot()
|
181 |
+
|
182 |
+
st.write("The prediction isn't good enough? Try to change settings in the model setup or increase the dataset length.")
|
183 |
+
st.write('Training took {} epochs, Mean Squared Error: {:.2e}'.format(result['epochs'], result['loss']))
|
184 |
+
#st.write('Training took {} epochs, Mean Squared Error {}, last loss {}'.format(result['epochs'], result['loss'], result['loss_last']))
|
button_style.css
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:388e988962ab53220113c1c3c41f470d8a3bde4e0cd43e0e45b1a51751a7816c
|
3 |
+
size 163
|
info.jpg
ADDED
![]() |
Git LFS Details
|
predict.py
ADDED
@@ -0,0 +1,152 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import tensorflow.keras as tf
|
2 |
+
import numpy as np
|
3 |
+
from sklearn.preprocessing import StandardScaler
|
4 |
+
|
5 |
+
verbose = 0
|
6 |
+
|
7 |
+
|
8 |
+
def predict_series(values, r1_nodes=5, r2_nodes=0, fc1_nodes=0, steps=20, use_lstm=True, *args, **kwargs):
|
9 |
+
|
10 |
+
train = np.array(values)
|
11 |
+
|
12 |
+
train_last_value = train[-1]
|
13 |
+
train = train[1:] - train[:-1]
|
14 |
+
sc = StandardScaler()
|
15 |
+
train = sc.fit_transform(train.reshape(-1, 1))
|
16 |
+
|
17 |
+
T = 25
|
18 |
+
X = []
|
19 |
+
Y = []
|
20 |
+
for t in range(len(train) - T):
|
21 |
+
x = train[t:t + T]
|
22 |
+
X.append(x)
|
23 |
+
Y.append(train[t + T])
|
24 |
+
|
25 |
+
X = np.array(X).reshape(-1, T, 1)
|
26 |
+
Y = np.array(Y)
|
27 |
+
|
28 |
+
nb_stats = 0
|
29 |
+
"""
|
30 |
+
X_temp = np.zeros(X.size + nb_stats * len(X)).reshape(-1, T + nb_stats)
|
31 |
+
|
32 |
+
step_size = 1 / (len(X) + steps)
|
33 |
+
|
34 |
+
|
35 |
+
def update_stats(row):
|
36 |
+
new_stat = row[T:]
|
37 |
+
new_stat[0] += step_size # number of sample
|
38 |
+
|
39 |
+
minimum = min(row[:T]) # minimum value, and when it occurred
|
40 |
+
if minimum < row[T + 1]:
|
41 |
+
new_stat[1], new_stat[2] = minimum, new_stat[0]
|
42 |
+
|
43 |
+
maximum = max(row[:T]) # maximum value, and when it occurred
|
44 |
+
if maximum > row[T + 3]:
|
45 |
+
new_stat[3], new_stat[4] = maximum, new_stat[0]
|
46 |
+
|
47 |
+
new_stat[5] = (row[T + 5] * row[T] + row[T - 1]) / (new_stat[0]) # rolling average
|
48 |
+
|
49 |
+
difference10 = row[T - 1] - row[T - 11] # the biggest difference within 10 items
|
50 |
+
if difference10 > row[T + 6]:
|
51 |
+
new_stat[6], new_stat[7] = difference10, new_stat[0]
|
52 |
+
if difference10 < row[T + 8]:
|
53 |
+
new_stat[8], new_stat[9] = difference10, new_stat[0]
|
54 |
+
|
55 |
+
abs_difference10 = abs(difference10) # the biggest absolute difference within 10 items
|
56 |
+
if abs_difference10 > row[T + 10]:
|
57 |
+
new_stat[10], new_stat[11] = abs_difference10, new_stat[0]
|
58 |
+
if abs_difference10 < row[T + 12]:
|
59 |
+
new_stat[12], new_stat[13] = abs_difference10, new_stat[0]
|
60 |
+
|
61 |
+
return new_stat
|
62 |
+
|
63 |
+
X_temp[0] = X[0] #np.append(X[0])#, [0, np.inf, 0, -np.inf, 0]) #, 0, -np.inf, 0, +np.inf, 0, 0, 0, np.inf, 0])
|
64 |
+
for i in range(1, len(X)):
|
65 |
+
X_temp[i] = np.append(X[i][:T], X_temp[i - 1][T:])
|
66 |
+
X_temp[i][T:] = update_stats(X_temp[i])
|
67 |
+
"""
|
68 |
+
#X = X_temp[1:].reshape(-1, T + nb_stats, 1)
|
69 |
+
#Y = Y[1:]
|
70 |
+
|
71 |
+
i = tf.layers.Input(shape=(T + nb_stats, 1))
|
72 |
+
|
73 |
+
if use_lstm:
|
74 |
+
rnn_layer = tf.layers.LSTM
|
75 |
+
else:
|
76 |
+
rnn_layer = tf.layers.GRU
|
77 |
+
|
78 |
+
if r2_nodes:
|
79 |
+
x = rnn_layer(r1_nodes, return_sequences=True)(i)
|
80 |
+
x = rnn_layer(r2_nodes)(x)
|
81 |
+
else:
|
82 |
+
x = rnn_layer(r1_nodes)(i)
|
83 |
+
if fc1_nodes:
|
84 |
+
x = tf.layers.Dense(fc1_nodes, activation='relu')(x)
|
85 |
+
x = tf.layers.Dense(1)(x)
|
86 |
+
model = tf.models.Model(i, x)
|
87 |
+
|
88 |
+
|
89 |
+
"""lr_schedule = tf.optimizers.schedules.ExponentialDecay(
|
90 |
+
initial_learning_rate=0.2,
|
91 |
+
decay_steps=10,
|
92 |
+
decay_rate=0.8)
|
93 |
+
optimizer = tf.optimizers.Ftrl(learning_rate=0.001, learning_rate_power=-0.1)"""
|
94 |
+
#for i in range(0, 500, 10):
|
95 |
+
#print('{}: {}'.format(i, lr_schedule(i)))
|
96 |
+
|
97 |
+
|
98 |
+
model.compile(
|
99 |
+
loss='mse', #tf.losses.LogCosh(),
|
100 |
+
optimizer=tf.optimizers.Adamax(lr=0.1) #LogCosh()'sgd'
|
101 |
+
)
|
102 |
+
|
103 |
+
callbacks = [tf.callbacks.EarlyStopping(patience=150, monitor='loss', restore_best_weights=True)]
|
104 |
+
|
105 |
+
r = model.fit(
|
106 |
+
X, Y,
|
107 |
+
epochs=500,
|
108 |
+
callbacks=callbacks,
|
109 |
+
verbose=verbose,
|
110 |
+
validation_split=0.0
|
111 |
+
)
|
112 |
+
pred = np.array([])
|
113 |
+
last_x = X[-1]
|
114 |
+
|
115 |
+
|
116 |
+
for _ in range(steps):
|
117 |
+
p = model.predict(last_x.reshape(1, -1, 1))[0, 0]
|
118 |
+
pred = np.append(pred, p)
|
119 |
+
#last_x[:T] = np.roll(last_x[:T], -1)
|
120 |
+
#last_x[T - 1] = p
|
121 |
+
#last_x[T:] = update_stats(last_x)
|
122 |
+
last_x = np.roll(last_x, -1)
|
123 |
+
last_x[-1] = p
|
124 |
+
|
125 |
+
pred = sc.inverse_transform(pred.reshape(-1, 1))
|
126 |
+
# pred = np.array(pred).astype('float64')
|
127 |
+
# pred = list(pred)
|
128 |
+
# logging.info(pred)
|
129 |
+
|
130 |
+
pred.reshape(-1)
|
131 |
+
pred[0] = train_last_value + pred[0]
|
132 |
+
|
133 |
+
for i in range(1, len(pred)):
|
134 |
+
pred[i] += pred[i-1]
|
135 |
+
|
136 |
+
|
137 |
+
result = {'result': list(pred.reshape(-1)), 'epochs': r.epoch[-1] + 1, 'loss': min(r.history['loss']), 'loss_last': r.history['loss'][-1]}
|
138 |
+
return result
|
139 |
+
|
140 |
+
|
141 |
+
if __name__ == "__main__":
|
142 |
+
from time import time
|
143 |
+
t1 = time()
|
144 |
+
verbose = 2
|
145 |
+
data = np.sin(np.arange(0.0, 28.0, 0.35)*2)
|
146 |
+
result = predict_series(data, steps=66, r1_nodes=14, r2_nodes=14, fc1_nodes=20)
|
147 |
+
print('exec time: {:8.3f}'.format(time()-t1))
|
148 |
+
#print(result['result'][:2])
|
149 |
+
print(print(result['epochs'], result['loss']))
|
150 |
+
import seaborn as sns
|
151 |
+
sns.lineplot(x=range(30), y=data[-30:], color='r')
|
152 |
+
sns.lineplot(x=range(30, 30+len(result['result'])), y=result['result'], color='b')
|
requirements.txt
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
streamlit==1.17.0
|
2 |
+
requests
|
3 |
+
numpy
|
4 |
+
seaborn
|
5 |
+
matplotlib
|
6 |
+
tensorflow
|
wait.gif
ADDED
![]() |
Git LFS Details
|