Julia Turc

@UCMUy_WsUIYbM_hKR9drb9dA - 59K subscribers

AI explainer videos from a former Google Research engineer, now startup founder. Expect American optimism with a solid dose of Eastern European cynicism. Find (free) supplemental materials (slide decks, paper reading lists, Colab notebooks) on Patreon 👉 www.patreon.com/c/JuliaTurc

Home Videos Live Playlists

Why are diffusion LLMs so fast?

Why are diffusion LLMs so fast? Julia Turc

14.2K views - 1 week ago

An image is worth NxN words | Diffusion Transformers (ViT, DiT, MMDiT)

An image is worth NxN words | Diffusion Transformers (ViT, DiT, MMDiT) Julia Turc

16.6K views - 1 week ago

Why are Transformers replacing CNNs?

Why are Transformers replacing CNNs? Julia Turc

166.5K views - 2 months ago

I asked them to show me their RAG pipeline...

I asked them to show me their RAG pipeline... Julia Turc

14.9K views - 2 months ago

David Petrou: Building social AI after 17 years at Google

David Petrou: Building social AI after 17 years at Google Julia Turc

3 months ago

Transformers & Diffusion LLMs: What's the connection?

Transformers & Diffusion LLMs: What's the connection? Julia Turc

34.1K views - 3 months ago

Text diffusion: A new paradigm for LLMs

Text diffusion: A new paradigm for LLMs Julia Turc

99.7K views - 4 months ago

Hierarchical Reasoning Model: Substance or Hype?

Hierarchical Reasoning Model: Substance or Hype? Julia Turc

25.5K views - 5 months ago

The physics behind diffusion models

The physics behind diffusion models Julia Turc

96.1K views - 5 months ago

Reverse-engineering GGUF | Post-Training Quantization

Reverse-engineering GGUF | Post-Training Quantization Julia Turc

47.8K views - 7 months ago

Training models with only 4 bits | Fully-Quantized Training

Training models with only 4 bits | Fully-Quantized Training Julia Turc

48.8K views - 7 months ago

The myth of 1-bit LLMs | Quantization-Aware Training

The myth of 1-bit LLMs | Quantization-Aware Training Julia Turc

86.4K views - 8 months ago

How LLMs survive in low precision | Quantization Fundamentals

How LLMs survive in low precision | Quantization Fundamentals Julia Turc

42.5K views - 8 months ago

Knowledge Distillation: How LLMs train each other

Knowledge Distillation: How LLMs train each other Julia Turc

59.9K views - 9 months ago

Mixture of Experts: How LLMs get bigger without getting slower

Mixture of Experts: How LLMs get bigger without getting slower Julia Turc

27.8K views - 9 months ago

Llama 4 Explained: Architecture, Long Context, and Native Multimodality

Llama 4 Explained: Architecture, Long Context, and Native Multimodality Julia Turc

15.8K views - 10 months ago

DeepSeek's GRPO (Group Relative Policy Optimization) | Reinforcement Learning for LLMs

DeepSeek's GRPO (Group Relative Policy Optimization) | Reinforcement Learning for LLMs Julia Turc

37.8K views - 10 months ago

Proximal Policy Optimization (PPO) for LLMs Explained Intuitively

Proximal Policy Optimization (PPO) for LLMs Explained Intuitively Julia Turc

44.8K views - 11 months ago

Tülu 3 from AI2: Full open-source fine-tuning recipe for LLMs

Tülu 3 from AI2: Full open-source fine-tuning recipe for LLMs Julia Turc

23.4K views - 11 months ago

8 Timeless tips for training LLMs | Become a better ML engineer

8 Timeless tips for training LLMs | Become a better ML engineer Julia Turc

11.4K views - 11 months ago

How does DeepSeek actually work? | Full technical review

How does DeepSeek actually work? | Full technical review Julia Turc

8.5K views - 1 year ago

Discover How LLMs Work by Dissecting Llama

Discover How LLMs Work by Dissecting Llama Julia Turc

13.1K views - 1 year ago