Skill

deep-learning

Deep Learning expertise from Google DeepMind - creators of AlphaGo, Transformers, and BERT. Master neural network architectures, training at scale, and production ML. Use when building ML models, training neural networks, or deploying AI systems. Learn from the team behind Google Search, YouTube recommendations, and Google Photos.

npx claudepluginhub duylinhdang1998/claude-template-agent --plugin vfm-agent-company

Tool Access

This skill uses the workspace's default tool permissions.

Preview

**Expert**: Dr. Emily Rodriguez (Google DeepMind, 10 years)

Supporting Assets

references/training.mdreferences/transformers.md

SKILL.md

Similar Skills

skill-lookup

159.9k

Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.

prompts.chat

prompt-lookup

159.9k

Searches prompts.chat for AI prompt templates by keyword or category, retrieves by ID with variable handling, and improves prompts via AI. Use for discovering or enhancing prompts.

prompts.chat

next-compile

139.2k

Checks Next.js compilation errors using a running Turbopack dev server after code edits. Fixes actionable issues before reporting complete. Replaces `next build`.

1 file

vercel-next-js-2

Stats

Parent Repo Stars0

Parent Repo Forks0

Last CommitApr 16, 2026

Actions

View Source View Plugin View on GitHub View README

Deep Learning - Google DeepMind Expertise

Expert: Dr. Emily Rodriguez (Google DeepMind, 10 years) Level: 10/10 - Google DeepMind researcher

Overview

Deep Learning at Google scale - neural networks powering billions of users across Search, YouTube, Photos, Assistant. Not toy models - production ML at the highest scale.

When to Use What

Computer Vision:

Image classification → CNNs - Object detection → Faster R-CNN, YOLO
Vision Transformers → ViT

Natural Language:

Text understanding → Transformers - BERT, T5, PaLM models
Fine-tuning for tasks

Sequence Modeling:

Time series → RNNs, LSTM - Audio/speech → Wav2Vec, Conformer

Core Workflow

Data preparation - Clean, augment, split datasets
Model selection - Choose architecture for task
Training - Use TPUs, distributed training 4. Evaluation - Rigorous metrics, validation
Optimization - Hyperparameter tuning, architecture search
Deployment - TensorFlow Serving, optimization 7. Monitoring - Track performance, detect drift

Google's Innovations

Transformers (2017): Attention mechanism revolutionized NLP BERT (2018): Bidirectional language understanding Vision Transformer (2020): Transformers for images PaLM (2022): 540B parameter language model Gemini (2023): Multimodal AI

Training at Scale

Google Infrastructure:

TPU v4/v5 pods (thousands of chips)
Distributed training strategies
Mixed precision (bfloat16)
Gradient accumulation

Production ML at Google

Powers:

Google Search: Billions of queries/day
YouTube: Video recommendations, 1B hours/day
Google Photos: 4 trillion photos, auto-tagging
Google Translate: 100+ languages
Google Assistant: Natural conversations

Best Practices

Start with pre-trained models - Transfer learning
Use TPUs for large models - Cost-effective training
Monitor training closely - TensorBoard, W&B
Validate rigorously - Multiple metrics, holdout sets
A/B test in production - Measure real impact

Related Skills

gcp-expert - Vertex AI, TPUs
nlp-expert - NLP applications
ml-infrastructure - MLOps

Last Updated: 2026-02-03 Expert: Dr. Emily Rodriguez (Google DeepMind, 10 years) - AlphaGo, Transformers