github-starred/TinyTorch
2
0
Fork 0
mirror of https://github.com/MLSysBook/TinyTorch.git synced 2026-03-12 02:09:16 -05:00
Code Issues 7 Packages Projects Releases Wiki Activity

Add live spinner to milestone training loops

Browse Source 
Use rich.live.Live to show real-time progress indicator during epoch training.
This gives visual feedback that code is running during potentially slow operations.
This commit is contained in:
Vijay Janapa Reddi
2025-11-22 15:31:48 -05:00
parent c77b05797f
commit 5c3695a797
3 changed files with 133 additions and 100 deletions
Download Patch File Download Diff File Expand all files Collapse all files

67
milestones/01_1957_perceptron/02_rosenblatt_trained.py
View File

@@ -40,6 +40,9 @@ from tinytorch import Tensor, Linear, Sigmoid, BinaryCrossEntropyLoss, SGD
from rich.console import Console
from rich.panel import Panel
from rich.table import Table
from rich.live import Live
from rich.spinner import Spinner
from rich.text import Text
from rich import box
console = Console()
@@ -107,40 +110,48 @@ def generate_data(n_samples=100, seed=None):
def train_perceptron(model, X, y, epochs=100, lr=0.1):
"""Train the perceptron using SGD."""
# Setup training components
loss_fn = BinaryCrossEntropyLoss()
optimizer = SGD(model.parameters(), lr=lr)
console.print("\n[bold cyan]🔥 Starting Training...[/bold cyan]\n")
history = {"loss": [], "accuracy": []}
for epoch in range(epochs):
# Forward pass
predictions = model(X)
loss = loss_fn(predictions, y)
# Backward pass
loss.backward()
# Update weights
optimizer.step()
optimizer.zero_grad()
# Calculate accuracy
pred_classes = (predictions.data > 0.5).astype(int)
accuracy = (pred_classes == y.data).mean()
history["loss"].append(loss.data.item())
history["accuracy"].append(accuracy)
# Print progress every 10 epochs
if (epoch + 1) % 10 == 0:
console.print(f"Epoch {epoch+1:3d}/{epochs} Loss: {loss.data:.4f} Accuracy: {accuracy:.1%}")
# Use Live display with spinner for real-time feedback
with Live(console=console, refresh_per_second=10) as live:
for epoch in range(epochs):
# Forward pass
predictions = model(X)
loss = loss_fn(predictions, y)
# Backward pass
loss.backward()
# Update weights
optimizer.step()
optimizer.zero_grad()
# Calculate accuracy
pred_classes = (predictions.data > 0.5).astype(int)
accuracy = (pred_classes == y.data).mean()
history["loss"].append(loss.data.item())
history["accuracy"].append(accuracy)
# Update spinner with current progress
spinner_text = Text()
spinner_text.append("⠋ ", style="cyan")
spinner_text.append(f"Epoch {epoch+1:3d}/{epochs} Loss: {loss.data:.4f} Accuracy: {accuracy:.1%}")
live.update(spinner_text)
# Print progress every 10 epochs
if (epoch + 1) % 10 == 0:
live.console.print(f"Epoch {epoch+1:3d}/{epochs} Loss: {loss.data:.4f} Accuracy: {accuracy:.1%}")
console.print("\n[bold green]✅ Training Complete![/bold green]\n")
return history

66
milestones/02_1969_xor/02_xor_solved.py
View File

@@ -64,6 +64,8 @@ import numpy as np
from rich.console import Console
from rich.table import Table
from rich.panel import Panel
from rich.live import Live
from rich.text import Text
from rich import box
# Add project root to path
@@ -160,40 +162,50 @@ class XORNetwork:
def train_network(model, X, y, epochs=500, lr=0.5):
"""
Train multi-layer network on XOR.
This WILL succeed - hidden layers solve the problem!
"""
loss_fn = BinaryCrossEntropyLoss()
optimizer = SGD(model.parameters(), lr=lr)
console.print("\n[bold cyan]🔥 Training Multi-Layer Network...[/bold cyan]")
console.print("[dim](This will work - hidden layers solve XOR!)[/dim]\n")
history = {"loss": [], "accuracy": []}
for epoch in range(epochs):
# Forward pass
predictions = model(X)
loss = loss_fn(predictions, y)
# Backward pass (through hidden layers!)
loss.backward()
# Update weights
optimizer.step()
optimizer.zero_grad()
# Calculate accuracy
pred_classes = (predictions.data > 0.5).astype(int)
accuracy = (pred_classes == y.data).mean()
history["loss"].append(loss.data.item())
history["accuracy"].append(accuracy)
# Print progress every 100 epochs
if (epoch + 1) % 100 == 0:
console.print(f"Epoch {epoch+1:3d}/{epochs} Loss: {loss.data:.4f} Accuracy: {accuracy:.1%}")
# Use Live display with spinner for real-time feedback
with Live(console=console, refresh_per_second=10) as live:
for epoch in range(epochs):
# Forward pass
predictions = model(X)
loss = loss_fn(predictions, y)
# Backward pass (through hidden layers!)
loss.backward()
# Update weights
optimizer.step()
optimizer.zero_grad()
# Calculate accuracy
pred_classes = (predictions.data > 0.5).astype(int)
accuracy = (pred_classes == y.data).mean()
history["loss"].append(loss.data.item())
history["accuracy"].append(accuracy)
# Update spinner with current progress
spinner_text = Text()
spinner_text.append("⠋ ", style="cyan")
spinner_text.append(f"Epoch {epoch+1:3d}/{epochs} Loss: {loss.data:.4f} Accuracy: {accuracy:.1%}")
live.update(spinner_text)
# Print progress every 100 epochs
if (epoch + 1) % 100 == 0:
live.console.print(f"Epoch {epoch+1:3d}/{epochs} Loss: {loss.data:.4f} Accuracy: {accuracy:.1%}")
console.print("\n[green]✅ Training Complete - XOR Solved![/green]")
return history

100
milestones/03_1986_mlp/01_rumelhart_tinydigits.py
View File

@@ -58,6 +58,8 @@ from tinytorch.data.loader import TensorDataset, DataLoader
from rich.console import Console
from rich.panel import Panel
from rich.table import Table
from rich.live import Live
from rich.text import Text
from rich import box
console = Console()
@@ -374,51 +376,59 @@ def train_mlp():
"train_accuracy": [],
"test_accuracy": []
}
for epoch in range(epochs):
epoch_loss = 0.0
batch_count = 0
for batch_images, batch_labels in train_loader:
# Forward pass
logits = model(batch_images)
loss = loss_fn(logits, batch_labels)
# Backward pass
loss.backward()
# Update weights
optimizer.step()
optimizer.zero_grad()
epoch_loss += loss.data
batch_count += 1
avg_loss = epoch_loss / batch_count
# Evaluate on both train and test to detect overfitting
train_acc, _ = evaluate_accuracy(model, train_images, train_labels)
test_acc, _ = evaluate_accuracy(model, test_images, test_labels)
history["train_loss"].append(avg_loss)
history["train_accuracy"].append(train_acc)
history["test_accuracy"].append(test_acc)
if initial_loss is None:
initial_loss = avg_loss
# Print progress every 5 epochs
if (epoch + 1) % 5 == 0:
gap = train_acc - test_acc
gap_indicator = "⚠️" if gap > 10 else "✓"
console.print(
f"Epoch {epoch+1:2d}/{epochs} "
f"Loss: {avg_loss:.4f} "
f"Train: {train_acc:.1f}% "
f"Test: {test_acc:.1f}% "
f"{gap_indicator} Gap: {gap:.1f}%"
)
# Use Live display with spinner for real-time feedback
with Live(console=console, refresh_per_second=10) as live:
for epoch in range(epochs):
epoch_loss = 0.0
batch_count = 0
for batch_images, batch_labels in train_loader:
# Forward pass
logits = model(batch_images)
loss = loss_fn(logits, batch_labels)
# Backward pass
loss.backward()
# Update weights
optimizer.step()
optimizer.zero_grad()
epoch_loss += loss.data
batch_count += 1
# Update spinner with current batch progress
spinner_text = Text()
spinner_text.append("⠋ ", style="cyan")
spinner_text.append(f"Epoch {epoch+1:2d}/{epochs} Batch {batch_count}/{len(train_loader)}")
live.update(spinner_text)
avg_loss = epoch_loss / batch_count
# Evaluate on both train and test to detect overfitting
train_acc, _ = evaluate_accuracy(model, train_images, train_labels)
test_acc, _ = evaluate_accuracy(model, test_images, test_labels)
history["train_loss"].append(avg_loss)
history["train_accuracy"].append(train_acc)
history["test_accuracy"].append(test_acc)
if initial_loss is None:
initial_loss = avg_loss
# Print progress every 5 epochs
if (epoch + 1) % 5 == 0:
gap = train_acc - test_acc
gap_indicator = "⚠️" if gap > 10 else "✓"
live.console.print(
f"Epoch {epoch+1:2d}/{epochs} "
f"Loss: {avg_loss:.4f} "
f"Train: {train_acc:.1f}% "
f"Test: {test_acc:.1f}% "
f"{gap_indicator} Gap: {gap:.1f}%"
)
console.print("\n[green]✅ Training Complete![/green]")
final_train_acc = history["train_accuracy"][-1]