Uploaded July 2020 | Updated November 2026, 4 days ago
CORRECTION: The score for BE is 6 and for BD is -1.
Restricted Boltzmann Machines (RBMs) are a type of neural network based on Hopfield models, created by Geoff Hinton, winner of the 2024 Physics Nobel Prize.
In this video you'll find a gentle introduction to RBMs and their training process, using a real-life example with people and pets.
Grokking Machine Learning Book: manning.com/books/grokking-machine-learning
40% discount promo code: serranoyt
Images downloaded from freepik.com
freepik.com/free-photos-vectors/banner
freepik.com/free-photos-vectors/dog
freepik.com/free-photos-vectors/people
freepik.com/free-photos-vectors/background
Introduction: (0:00)
Mystery: (0:17)
Scores: (4:39)
Probabilities: (7:30)
Training (11:09)
Contrastive Divergence: (13:37)
Small Problem: (15:33)
Gibbs Sampling: (16:33)
Updating Weights: (20:56)
Sampling Problems: (22:58)
Independent Sampling: (24:27)
Picking Random Samples with Conditions: (28:30)
Picking Completely Random Samples: (31:05)
Summary: (35:03)
Conclusion: (35:57)
CORRECTION: The score for BE is 6 and for BD is -1.
Restricted Boltzmann Machines (RBMs) are a type of neural network based on Hopfield models, created by Geoff Hinton, winner of the 2024 Physics Nobel Prize.
In this video you'll find a gentle introduction to RBMs and their training process, using a real-life example with people and pets.
Grokking Machine Learning Book: manning.com/books/grokking-machine-learning
40% discount promo code: serranoyt
Images downloaded from freepik.com
freepik.com/free-photos-vectors/banner
freepik.com/free-photos-vectors/dog
freepik.com/free-photos-vectors/people
freepik.com/free-photos-vectors/background
Introduction: (0:00)
Mystery: (0:17)
Scores: (4:39)
Probabilities: (7:30)
Training (11:09)
Contrastive Divergence: (13:37)
Small Problem: (15:33)
Gibbs Sampling: (16:33)
Updating Weights: (20:56)
Sampling Problems: (22:58)
Independent Sampling: (24:27)
Picking Random Samples with Conditions: (28:30)
Picking Completely Random Samples: (31:05)
Summary: (35:03)
Conclusion: (35:57)
![Low Rank Adaptation (LoRA) - How to search for a needle in a much smaller haystack
When you want to fine-tune a model with many parameters and you want to do it fast and effectively, Low Rank Adaptation (LoRA) is there to help you.
With LoRA, you trust that the base model is in a good state, and so you assume there may already be many dependencies among the parameters. Thus, you look for better parameters in a much smaller dimensional space.
Learn all about LoRA in this friendly video!
CORRECTION: (thanks @ryanhewitt9902!) At 30:14 the vector on the right should be [pa, qa, ra, sa, ta], instead of [p/a, q/a, r/a, s/a, t/a]. Low Rank Adaptation (LoRA) - How to search for a needle in a much smaller haystack](https://i.ytimg.com/vi/Gy9jrVQTY4Q/mqdefault.jpg)









