brycewang-stanford/Auto-Empirical-Research-Skills / skills/33-Galaxy-Dawn-claude-scholar/skills/architecture-design
architecture-design
VerifiedUse only when creating new registrable ML components that require Factory or Registry patterns.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 9.0 KB |
| examples/augmentation_example.py | script | 3.1 KB |
| examples/config_example.yaml | config | 2.3 KB |
| examples/custom_dataset.py | script | 1.5 KB |
| examples/custom_model.py | script | 6.1 KB |
| examples/pipeline_example.shexecutable | script | 5.0 KB |
| references/auto_import.md | markdown | 3.4 KB |
| references/code_style.md | markdown | 5.2 KB |
| references/factory_pattern.md | markdown | 1.4 KB |
| references/registry_pattern.md | markdown | 2.5 KB |
| references/structure.md | markdown | 4.8 KB |
examples/augmentation_example.py
View on GitHub"""
Data Augmentation Example
Demonstrates how to create a custom data augmentation function
following the architecture design pattern.
"""
import torch
from typing import Dict
from src.data_module.augmentation import register_augmentation
@register_augmentation("time_shift")
def time_shift(signal: torch.Tensor, max_shift: int = 10) -> torch.Tensor:
"""Randomly shift signal in time.
Args:
signal: Input signal tensor of shape (channels, time_steps)
max_shift: Maximum number of steps to shift
Returns:
Shifted signal tensor
"""
shift = torch.randint(-max_shift, max_shift + 1, (1,)).item()
return torch.roll(signal, shifts=shift, dims=-1)
@register_augmentation("amplitude_scale")
def amplitude_scale(
signal: torch.Tensor,
min_scale: float = 0.8,
max_scale: float = 1.2
) -> torch.Tensor:
"""Randomly scale signal amplitude.
Args:
signal: Input signal tensor of shape (channels, time_steps)
min_scale: Minimum scaling factor
max_scale: Maximum scaling factor
Returns:
Scaled signal tensor
"""
scale = torch.empty(1).uniform_(min_scale, max_scale).item()
return signal * scale
@register_augmentation("gaussian_noise")
def add_gaussian_noise(
signal: torch.Tensor,
mean: float = 0.0,
std: float = 0.1
) -> torch.Tensor:
"""Add Gaussian noise to signal.
Args:
signal: Input signal tensor of shape (channels, time_steps)
mean: Mean of Gaussian noise
std: Standard deviation of Gaussian noise
Returns:
Signal with added noise
"""
noise = torch.randn_like(signal) * std + mean
return signal + noise
# Example: Composed augmentation
@register_augmentation("composed")
def composed_augmentation(signal: torch.Tensor, cfg) -> torch.Tensor:
"""Apply multiple augmentations in sequence.
Args:
signal: Input signal tensor
cfg: Configuration object with augmentation parameters
Returns:
Augmented signal tensor
"""
# Apply each augmentation based on config
if cfg.augmentation.time_shift:
signal = time_shift(signal, cfg.augmentation.max_shift)
if cfg.augmentation.amplitude_scale:
signal = amplitude_scale(
signal,
cfg.augmentation.min_scale,
cfg.augmentation.max_scale
)
if cfg.augmentation.gaussian_noise:
signal = add_gaussian_noise(
signal,
cfg.augmentation.noise_mean,
cfg.augmentation.noise_std
)
return signal
# Usage in dataset class
class AugmentedDataset:
"""Example dataset with augmentation support."""
def __init__(self, cfg):
self.cfg = cfg
self.augmentation_fn = AugmentationFactory(cfg.augmentation.name)
def __getitem__(self, idx: int) -> Dict[str, torch.Tensor]:
# Load signal
signal = self.load_signal(idx)
# Apply augmentation (training mode only)
if self.training and self.augmentation_fn:
signal = self.augmentation_fn(signal, self.cfg)
return {"signal": signal, "label": self.labels[idx]}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| brainstormingobra/superpowersYou MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation). | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| brand-guidelinesanthropics/skillsApplies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| frontend-designanthropics/skillsGuidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| mcp-builderanthropics/skillsGuide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK). | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |