Deep Learning with JAX

This book provides a practical guide to building and optimizing deep learning models using the JAX library. Readers will learn to leverage its core capabilities such as automatic differentiation, JIT compilation, auto-vectorization, and parallelization for high-performance computing. The content also covers essential topics like array manipulation, random number generation, handling complex data structures, and integrating with higher-level neural network libraries within the JAX ecosystem to develop efficient and scalable machine learning applications.

This page is free — overview and chapter list only.
The complete book body is sold separately.

Full book: $2.99 USDC via x402 ·
Per chapter: $0.25 USDC

Buy / open complete book (HTML)
· Complete book Markdown (.md)

Agents: start with free /library/discovery.json, sample free teaser chapters,
then pay for individual chapters or the complete book URL above.
Append .md to any content URL for Markdown with YAML front matter.

Chapters

  1. Deep Learning With Jax (Free teaser)
  2. About The Cover Illustration ($0.25)
  3. First Steps ($0.25)
  4. When And Why To Use Jax ($0.25)
  5. Listing 21 Loading The Dataset Import Tensorflow As Tf Impor ($0.25)
  6. Listing 22 Showing Samples From The Dataset ($0.25)
  7. Gradient Descent Procedure ($0.25)
  8. 263 Gradient Update Step ($0.25)
  9. Listing 215 Saving And Loading Model Parameters ($0.25)
  10. Applying The Filter Kernel To An Image ($0.25)
  11. Jax ($0.25)
  12. Jax 2 ($0.25)
  13. Listing 410 Differentiating With Respect To Dicts ($0.25)
  14. Returning Auxiliary Data From A Function ($0.25)
  15. Listing 411 Returning Auxiliary Data From A Function ($0.25)
  16. Obtaining Both The Gradient And Value Of The Function ($0.25)
  17. Per Sample Gradients ($0.25)
  18. Stopping Gradients ($0.25)
  19. Listing 413 Stopping Gradient Flow ($0.25)
  20. Listing 414 Finding Higher Order Derivatives ($0.25)
  21. Listing 415 Drawing Higher Order Derivatives ($0.25)
  22. Jacobian Matrix ($0.25)
  23. Hessian Matrix ($0.25)
  24. Listing 417 Calculating The Hessian Of A Function ($0.25)
  25. 432 Forward Mode And Jvp ($0.25)
  26. Listing 418 Checking Manual Forward Mode Calculations ($0.25)
  27. Directional Derivative And Jvp ($0.25)
  28. Listing 419 Calculating Jvp ($0.25)
  29. Listing 420 Recovering Jacobian Columns With Jvp ($0.25)
  30. Listing 421 Checking Manual Forward Mode Calculations With J ($0.25)
  31. 433 Reverse Mode And Vjp ($0.25)
  32. Reverse Mode Generalization And Vjp ($0.25)
  33. Listing 424 Recovering Jacobian Rows With Vjp ($0.25)
  34. Listing 51 Selu Activation Function ($0.25)
  35. 511 Using Jit Compilation ($0.25)
  36. Jit And Aot ($0.25)
  37. Compiling And Running On Specific Hardware ($0.25)
  38. Using Static Arguments ($0.25)
  39. Listing 58 Using Functoolspartial With Jit As A Decorator ($0.25)
  40. Optimization Related Arguments ($0.25)
  41. Listing 59 Compiling An Impure Function ($0.25)
  42. 521 Jaxpr An Intermediate Representation For Jax Programs ($0.25)
  43. Listing 510 Using Jaxmakejaxpr ($0.25)
  44. Tracing ($0.25)
  45. Listing 516 Replacing For Loop With A Structured Control Flo ($0.25)
  46. Listing 517 Replacing An If Statement With A Structured Cont ($0.25)
  47. Xla Origins And Architecture ($0.25)
  48. Openxla ($0.25)
  49. Listing 714 Using The Inaxes Parameter For Broadcasting ($0.25)
  50. Communicating Between Processes ($0.25)
  51. Listing 728 Data Tensor Shapes ($0.25)
  52. In The Final Step We Can Run Our Program ($0.25)
  53. Listing 89 Different Devices ($0.25)
  54. Listing 92 Samples From The Dataset ($0.25)
  55. Listing 99 Generating Random Values In Numpy From Numpy Impo ($0.25)
  56. Listing 912 Looking At The Prng State For The Legacy Approac ($0.25)
  57. Listing 915 Looking At The Key Key Randomprngkey42 Typekey J ($0.25)
  58. Listing 917 Generating 100 Keys Key Randomprngkey42 Key Subk ($0.25)
  59. 121 Deep Learning Ecosystem ($0.25)
  60. C1 Setting Up A Cloud Tpu Project ($0.25)
  61. C21 Running And Deleting A Cloud Tpu Node ($0.25)
  62. C22 Preparing A Cloud Tpu Node ($0.25)
  63. C23 Connecting To A Cloud Tpu Node From A Colab Notebook ($0.25)
  64. C3 Resources ($0.25)
  65. D11 Working With Named Axes ($0.25)
  66. Named Axis Programming And Axisname In Vmappmap ($0.25)
  67. Einsum ($0.25)
  68. D12 Parallelism And Hardware Meshes ($0.25)
  69. D21 Basics Of Pjit ($0.25)
  70. Listing D13 Adding Input Partitioning For Both Arguments ($0.25)
  71. D22 Mlp Example With Pjit ($0.25)
  72. Related Manning Titles ($0.25)
  73. Deep Learning With Jax 2 ($0.25)
  74. Whats Inside ($0.25)