File size: 4,575 Bytes
e5aa730
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
182f1c9
e5aa730
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
import streamlit as st
import torch
import numpy as np
import plotly.graph_objs as go

st.markdown("""
    <style>
    .stApp {
        background-color: #f9f9f9;
        font-family: 'Segoe UI', sans-serif;
    }
    h1 {
        text-align: center;
        color: #2C3E50;
        font-size: 38px !important;
        font-weight: bold;
        margin-bottom: 20px;
    }
    .stTextInput > div > div > input {
        border: 2px solid #3498DB;
        border-radius: 8px;
        padding: 8px;
    }
    div.stButton > button {
        background-color: #3498DB;
        color: white;
        border-radius: 10px;
        padding: 10px 24px;
        font-size: 16px;
        border: none;
        transition: 0.3s;
    }
    div.stButton > button:hover {
        background-color: #2980B9;
        transform: scale(1.05);
    }
    .stAlert {
        border-radius: 8px;
    }
    .block-container {
        padding-top: 2rem;
        padding-bottom: 2rem;
        max-width: 1200px;
    }
    </style>
""", unsafe_allow_html=True)

st.title("Gradient Descent Visualizer")


func_input = st.text_input("Enter Function of x", "x**2")
start_point = float(st.text_input("Starting Point", "2"))
learning_rate = float(st.text_input("Learning Rate", "0.01"))
num_iterations = int(st.text_input("Number of Iterations", "10"))

def make_function(expr: str):
    """Dynamically create a function in torch"""
    def func(x):
        return eval(expr, {"x": x, "torch": torch})
    return func

if st.button("Set Up") or 'func' not in st.session_state or 'points' not in st.session_state:
    try:
        func = make_function(func_input)

        st.session_state.func = func
        st.session_state.points = [start_point]
        st.session_state.step = 0
        st.success("Function Set Up Successfully with PyTorch!")
    except Exception as e:
        st.error(f"Error setting up function: {e}")

def gradient_step(x_val, func, lr):
    x = torch.tensor([x_val], dtype=torch.float32, requires_grad=True)
    y = func(x)
    y.backward()
    grad = x.grad.item()
    new_x = x_val - lr * grad
    return new_x, grad

if 'func' in st.session_state:
    if st.button("Next Iteration"):
        try:
            x_old = float(st.session_state.points[-1])
            x_new, grad_val = gradient_step(x_old, st.session_state.func, learning_rate)
            st.session_state.points.append(x_new)
            st.session_state.step += 1
            st.success(f"Iteration {st.session_state.step} Complete! (grad={grad_val:.6f})")
        except Exception as e:
            st.error(f"Error in iteration: {e}")

    if st.button("Run Iterations"):
        try:
            for i in range(num_iterations):
                x_old = float(st.session_state.points[-1])
                x_new, grad_val = gradient_step(x_old, st.session_state.func, learning_rate)
                st.session_state.points.append(x_new)
                st.session_state.step += 1
            st.success(f"Ran {st.session_state.step} Iterations in total")
        except Exception as e:
            st.error(f"Error in multiple iterations: {e}")


if 'func' in st.session_state and len(st.session_state.points) > 0:
    try:
        x_val = np.linspace(-10, 10, 500)
        x_torch = torch.tensor(x_val, dtype=torch.float32)
        y_val = st.session_state.func(x_torch).detach().numpy()

        iter_points = np.array(st.session_state.points)
        iter_torch = torch.tensor(iter_points, dtype=torch.float32)
        iter_y = st.session_state.func(iter_torch).detach().numpy()

        trace1 = go.Scatter(x=x_val, y=y_val, mode="lines", name="Function", line=dict(color="blue"))
        trace2 = go.Scatter(x=iter_points, y=iter_y, mode="markers+lines", 
                            name="Gradient Descent Path", marker=dict(color="red"))
        trace3 = go.Scatter(x=[iter_points[-1]], y=[iter_y[-1]], mode='markers+text', 
                            marker=dict(color='green', size=15), 
                            text=[f"{iter_points[-1]:.6f}"], textposition="top center", 
                            name="Current Point")

        layout = go.Layout(
            title=f"Iteration {st.session_state.step}",
            xaxis=dict(title="x - axis"),
            yaxis=dict(title="y - axis"),
            width=1000,
            height=600
        )

        fig = go.Figure(data=[trace1, trace2, trace3], layout=layout)
        st.plotly_chart(fig, use_container_width=True)
        st.success(f"Current Point = {iter_points[-1]}")
    except Exception as e:
        st.error(f"Plot error: {e}")