58:04Attention is all you need (Transformer) - Model explanation (including math), Inference and Training
A 58-minute video from Umar Jamil's catalogue; the title names its topic.
Umar Jamil
92.6K subscribers on YouTube
Shorts, livestreams, trailers and promos are never listed. Unless a curator approved this channel in full, only individually picked videos appear.
58:04
54:52BERT explained: Training, Inference, BERT vs GPT/LLamA, Fine tuning, [CLS] token
A 55-minute video from Umar Jamil's catalogue; the title names its topic.
19:34:45Building a distributed training framework from first principles
A 1175-minute video from Umar Jamil's catalogue; the title names its topic.
14:01CLIP - Paper explanation (training and inference)
A 14-minute video from Umar Jamil's catalogue; the title names its topic.
3:04:11Coding LLaMA 2 from scratch in PyTorch - KV Cache, Grouped Query Attention, Rotary PE, RMSNorm
A 184-minute video from Umar Jamil's catalogue; the title names its topic.
5:03:32Coding Stable Diffusion from scratch in PyTorch
A 304-minute video from Umar Jamil's catalogue; the title names its topic.
5:46:05Coding a Multimodal (Vision) Language Model from scratch in PyTorch with full explanation
A 346-minute video from Umar Jamil's catalogue; the title names its topic.
2:59:24Coding a Transformer from scratch on PyTorch, with full explanation, training and inference.
A 179-minute video from Umar Jamil's catalogue; the title names its topic.
48:46Direct Preference Optimization (DPO) explained: Bradley-Terry model, log probabilities, math
A 49-minute video from Umar Jamil's catalogue; the title names its topic.
1:12:53Distributed Data Parallel (DDP) with PyTorch: complete tutorial with cloud infrastructure and code
A 73-minute video from Umar Jamil's catalogue; the title names its topic.
7:38:18Flash Attention derived and coded from first principles with Triton (Python)
A 458-minute video from Umar Jamil's catalogue; the title names its topic.
21:12How diffusion models work - explanation and code!
A 21-minute video from Umar Jamil: “How diffusion models work - explanation and code!”.
1:15:39Kolmogorov-Arnold Networks: MLP vs KAN, Math, B-Splines, Universal Approximation Theorem
A 76-minute video from Umar Jamil's catalogue; the title names its topic.
1:10:55LLaMA explained: KV-Cache, Rotary Positional Embedding, RMS Norm, Grouped Query Attention, SwiGLU
A 71-minute video from Umar Jamil's catalogue; the title names its topic.
26:55LoRA: Low-Rank Adaptation of Large Language Models - Explained visually + PyTorch code from scratch
A 27-minute video from Umar Jamil's catalogue; the title names its topic.
29:58LongNet: Scaling Transformers to 1,000,000,000 tokens: Python Code + Explanation
A 30-minute video from Umar Jamil's catalogue; the title names its topic.
1:00:15ML Interpretability: feature visualization, adversarial example, interp. for language models
A 60-minute video from Umar Jamil's catalogue; the title names its topic.
1:14:29Mamba and S4 Explained: Architecture, Parallel Scan, Kernel Fusion, Recurrent, Convolution, Math
A 74-minute video from Umar Jamil's catalogue; the title names its topic.
1:26:21Mistral / Mixtral Explained: Sliding Window Attention, Sparse Mixture of Experts, Rolling Buffer
A 86-minute video from Umar Jamil's catalogue; the title names its topic.
50:55Quantization explained with PyTorch - Post-Training Quantization, Quantization-Aware Training
A 51-minute video from Umar Jamil's catalogue; the title names its topic.
2:15:13Reinforcement Learning from Human Feedback explained with math derivations and the PyTorch code.
A 135-minute video from Umar Jamil's catalogue; the title names its topic.
49:24Retrieval Augmented Generation (RAG) Explained: Embedding, Sentence BERT, Vector Database (HNSW)
A 49-minute video from Umar Jamil's catalogue; the title names its topic.
42:53Segment Anything - Model explanation with code
A 43-minute video from Umar Jamil's catalogue; the title names its topic.
27:12Variational Autoencoder - Model, ELBO, loss function and maths explained easily!
A 27-minute video from Umar Jamil: “Variational Autoencoder - Model, ELBO, loss function and maths explained easily!”.
6:58Wav2Lip (generate talking avatar videos) - Paper reading and explanation
A 7-minute video from Umar Jamil's catalogue; the title names its topic.