ByteChai@bytechai

Evening class ยท 20:00 IST

AI from scratch

100 lessons in 8 modules, posted one per evening and strictly in order. Each lesson builds on the one before it, so start at lesson 1 and keep going.

0 / 100 postedNext up: lesson 1, What AI actually is
  1. Lesson 1: What AI actually is, coming soon
  2. Lesson 2: Rules vs learning, coming soon
  3. Lesson 3: Narrow AI vs general AI, coming soon
  4. Lesson 4: The Turing test (1950), coming soon
  5. Lesson 5: Dartmouth 1956: AI gets a name, coming soon
  6. Lesson 6: Expert systems and the AI winters, coming soon
  7. Lesson 7: Why AI took off after 2010, coming soon
  8. Lesson 8: AI, ML, deep learning, LLMs, coming soon
  9. Lesson 9: Data, features and labels, coming soon
  10. Lesson 10: Training vs inference, coming soon
  11. Lesson 11: Supervised learning, coming soon
  12. Lesson 12: Unsupervised learning, coming soon
  13. Lesson 13: Reinforcement learning, coming soon
  14. Lesson 14: Linear regression, coming soon
  15. Lesson 15: Loss functions, coming soon
  16. Lesson 16: Gradient descent, coming soon
  17. Lesson 17: Learning rate, coming soon
  18. Lesson 18: Classification and logistic regression, coming soon
  19. Lesson 19: Overfitting vs underfitting, coming soon
  20. Lesson 20: Train, validation and test splits, coming soon
  21. Lesson 21: Accuracy, precision and recall, coming soon
  22. Lesson 22: Decision trees and random forests, coming soon
  23. Lesson 23: The artificial neuron, coming soon
  24. Lesson 24: Layers and deep networks, coming soon
  25. Lesson 25: Activation functions, coming soon
  26. Lesson 26: Forward pass, coming soon
  27. Lesson 27: Backpropagation, coming soon
  28. Lesson 28: Epochs and batches, coming soon
  29. Lesson 29: Why GPUs matter, coming soon
  30. Lesson 30: ImageNet and AlexNet (2012), coming soon
  31. Lesson 31: Convolutional networks for images, coming soon
  32. Lesson 32: Recurrent networks for sequences, coming soon
  33. Lesson 33: Vanishing gradients and LSTMs, coming soon
  34. Lesson 34: Dropout and regularisation, coming soon
  35. Lesson 35: Transfer learning, coming soon
  36. Lesson 36: GANs: two networks competing, coming soon
  37. Lesson 37: How computers read text, coming soon
  38. Lesson 38: Tokens and tokenization, coming soon
  39. Lesson 39: Bag of words and TF-IDF, coming soon
  40. Lesson 40: Word embeddings, coming soon
  41. Lesson 41: word2vec: king - man + woman, coming soon
  42. Lesson 42: Cosine similarity, coming soon
  43. Lesson 43: Sentence embeddings, coming soon
  44. Lesson 44: Semantic search, coming soon
  45. Lesson 45: Vector databases, coming soon
  46. Lesson 46: Seq2seq and machine translation, coming soon
  47. Lesson 47: The bottleneck problem, coming soon
  48. Lesson 48: Attention, the first version, coming soon
  49. Lesson 49: Attention Is All You Need (2017), coming soon
  50. Lesson 50: Self-attention, intuitively, coming soon
  51. Lesson 51: Queries, keys and values, coming soon
  52. Lesson 52: Multi-head attention, coming soon
  53. Lesson 53: Positional encoding, coming soon
  54. Lesson 54: Encoder vs decoder models, coming soon
  55. Lesson 55: Why transformers scale, coming soon
  56. Lesson 56: Context windows, coming soon
  57. Lesson 57: Next-token prediction, coming soon
  58. Lesson 58: Temperature and sampling, coming soon
  59. Lesson 59: Pre-training, coming soon
  60. Lesson 60: Scaling laws, coming soon
  61. Lesson 61: From GPT-1 to GPT-3, coming soon
  62. Lesson 62: Emergent abilities, coming soon
  63. Lesson 63: Instruction tuning, coming soon
  64. Lesson 64: RLHF, coming soon
  65. Lesson 65: ChatGPT, November 2022, coming soon
  66. Lesson 66: Hallucinations, coming soon
  67. Lesson 67: Parameters vs tokens, coming soon
  68. Lesson 68: Open vs closed models, coming soon
  69. Lesson 69: Quantization, coming soon
  70. Lesson 70: Running models locally, coming soon
  71. Lesson 71: Multimodal models, coming soon
  72. Lesson 72: Reasoning models, coming soon
  73. Lesson 73: Benchmarks and their limits, coming soon
  74. Lesson 74: API pricing and tokens, coming soon
  75. Lesson 75: Prompt engineering basics, coming soon
  76. Lesson 76: System prompts, coming soon
  77. Lesson 77: Few-shot examples, coming soon
  78. Lesson 78: Chain-of-thought prompting, coming soon
  79. Lesson 79: Structured outputs, coming soon
  80. Lesson 80: Tool calling, coming soon
  81. Lesson 81: RAG explained, coming soon
  82. Lesson 82: Chunking documents, coming soon
  83. Lesson 83: Reranking, coming soon
  84. Lesson 84: Evals, coming soon
  85. Lesson 85: Fine-tuning vs RAG vs prompting, coming soon
  86. Lesson 86: LoRA fine-tuning, coming soon
  87. Lesson 87: Prompt injection, coming soon
  88. Lesson 88: Prompt caching and latency, coming soon
  89. Lesson 89: What an AI agent is, coming soon
  90. Lesson 90: The agent loop, coming soon
  91. Lesson 91: Memory for agents, coming soon
  92. Lesson 92: MCP: Model Context Protocol, coming soon
  93. Lesson 93: Multi-agent systems, coming soon
  94. Lesson 94: AI coding assistants, coming soon
  95. Lesson 95: Agentic coding workflows, coming soon
  96. Lesson 96: Computer-use agents, coming soon
  97. Lesson 97: Guardrails and human-in-the-loop, coming soon
  98. Lesson 98: Deploying AI features, coming soon
  99. Lesson 99: Responsible AI for developers, coming soon
  100. Lesson 100: What comes next, coming soon

Module 01

What AI actually is

0/8 posted

  1. 002Rules vs learningUpcoming
  2. 003Narrow AI vs general AIUpcoming
  3. 004The Turing test (1950)Upcoming
  4. 005Dartmouth 1956: AI gets a nameUpcoming
  5. 006Expert systems and the AI wintersUpcoming
  6. 007Why AI took off after 2010Upcoming
  7. 008AI, ML, deep learning, LLMsUpcoming

Module 02

Machine learning basics

0/14 posted

  1. 009Data, features and labelsUpcoming
  2. 010Training vs inferenceUpcoming
  3. 011Supervised learningUpcoming
  4. 012Unsupervised learningUpcoming
  5. 013Reinforcement learningUpcoming
  6. 014Linear regressionUpcoming
  7. 015Loss functionsUpcoming
  8. 016Gradient descentUpcoming
  9. 017Learning rateUpcoming
  10. 018Classification and logistic regressionUpcoming
  11. 019Overfitting vs underfittingUpcoming
  12. 020Train, validation and test splitsUpcoming
  13. 021Accuracy, precision and recallUpcoming
  14. 022Decision trees and random forestsUpcoming

Module 03

Neural networks

0/14 posted

  1. 023The artificial neuronUpcoming
  2. 024Layers and deep networksUpcoming
  3. 025Activation functionsUpcoming
  4. 026Forward passUpcoming
  5. 027BackpropagationUpcoming
  6. 028Epochs and batchesUpcoming
  7. 029Why GPUs matterUpcoming
  8. 030ImageNet and AlexNet (2012)Upcoming
  9. 031Convolutional networks for imagesUpcoming
  10. 032Recurrent networks for sequencesUpcoming
  11. 033Vanishing gradients and LSTMsUpcoming
  12. 034Dropout and regularisationUpcoming
  13. 035Transfer learningUpcoming
  14. 036GANs: two networks competingUpcoming

Module 04

Language and embeddings

0/12 posted

  1. 037How computers read textUpcoming
  2. 038Tokens and tokenizationUpcoming
  3. 039Bag of words and TF-IDFUpcoming
  4. 040Word embeddingsUpcoming
  5. 041word2vec: king - man + womanUpcoming
  6. 042Cosine similarityUpcoming
  7. 043Sentence embeddingsUpcoming
  8. 044Semantic searchUpcoming
  9. 045Vector databasesUpcoming
  10. 046Seq2seq and machine translationUpcoming
  11. 047The bottleneck problemUpcoming
  12. 048Attention, the first versionUpcoming

Module 05

Transformers

0/12 posted

  1. 049Attention Is All You Need (2017)Upcoming
  2. 050Self-attention, intuitivelyUpcoming
  3. 051Queries, keys and valuesUpcoming
  4. 052Multi-head attentionUpcoming
  5. 053Positional encodingUpcoming
  6. 054Encoder vs decoder modelsUpcoming
  7. 055Why transformers scaleUpcoming
  8. 056Context windowsUpcoming
  9. 057Next-token predictionUpcoming
  10. 058Temperature and samplingUpcoming
  11. 059Pre-trainingUpcoming
  12. 060Scaling lawsUpcoming

Module 06

Large language models

0/14 posted

  1. 061From GPT-1 to GPT-3Upcoming
  2. 062Emergent abilitiesUpcoming
  3. 063Instruction tuningUpcoming
  4. 064RLHFUpcoming
  5. 065ChatGPT, November 2022Upcoming
  6. 066HallucinationsUpcoming
  7. 067Parameters vs tokensUpcoming
  8. 068Open vs closed modelsUpcoming
  9. 069QuantizationUpcoming
  10. 070Running models locallyUpcoming
  11. 071Multimodal modelsUpcoming
  12. 072Reasoning modelsUpcoming
  13. 073Benchmarks and their limitsUpcoming
  14. 074API pricing and tokensUpcoming

Module 07

Building with LLMs

0/14 posted

  1. 075Prompt engineering basicsUpcoming
  2. 076System promptsUpcoming
  3. 077Few-shot examplesUpcoming
  4. 078Chain-of-thought promptingUpcoming
  5. 079Structured outputsUpcoming
  6. 080Tool callingUpcoming
  7. 081RAG explainedUpcoming
  8. 082Chunking documentsUpcoming
  9. 083RerankingUpcoming
  10. 084EvalsUpcoming
  11. 085Fine-tuning vs RAG vs promptingUpcoming
  12. 086LoRA fine-tuningUpcoming
  13. 087Prompt injectionUpcoming
  14. 088Prompt caching and latencyUpcoming

Module 08

Agents and the AI development era

0/12 posted

  1. 089What an AI agent isUpcoming
  2. 090The agent loopUpcoming
  3. 091Memory for agentsUpcoming
  4. 092MCP: Model Context ProtocolUpcoming
  5. 093Multi-agent systemsUpcoming
  6. 094AI coding assistantsUpcoming
  7. 095Agentic coding workflowsUpcoming
  8. 096Computer-use agentsUpcoming
  9. 097Guardrails and human-in-the-loopUpcoming
  10. 098Deploying AI featuresUpcoming
  11. 099Responsible AI for developersUpcoming
  12. 100What comes nextUpcoming