Это пробный урок. Оформите подписку, чтобы получить доступ ко всем материалам курса. Премиум

  1. Урок 1. 00:01:36
    AI Engineering Bootcamp: Learn AWS SageMaker with Patrik Szepesi
  2. Урок 2. 00:08:43
    Course Introduction Part 1: Build, Train, and Deploy Models with AWS SageMaker
  3. Урок 3. 00:05:20
    Course Introduction Part 2: Fine-Tuning LLMs with QLoRA, AWS, and Open Source
  4. Урок 4. 00:04:32
    Setting Up Our AWS Account
  5. Урок 5. 00:07:40
    Set Up IAM Roles + Best Practices
  6. Урок 6. 00:07:02
    AWS Security Best Practices
  7. Урок 7. 00:02:23
    Set Up AWS SageMaker Domain
  8. Урок 8. 00:00:43
    UI Domain Change
  9. Урок 9. 00:02:41
    Sagemaker Domain Creation Update Part 1
  10. Урок 10. 00:03:07
    Sagemaker Domain Creation Update Part 2
  11. Урок 11. 00:11:59
    Sagemaker Notebooks Update
  12. Урок 12. 00:05:09
    Setting Up SageMaker Environment
  13. Урок 13. 00:08:45
    SageMaker Studio and Pricing
  14. Урок 14. 00:07:36
    Quota Increase
  15. Урок 15. 00:06:09
    Setup: SageMaker Server + PyTorch
  16. Урок 16. 00:18:35
    HuggingFace Models, Sentiment Analysis, and AutoScaling
  17. Урок 17. 00:06:04
    Get Dataset for Multiclass Text Classification
  18. Урок 18. 00:03:53
    Creating Our AWS S3 Bucket
  19. Урок 19. 00:01:27
    Uploading Our Training Data to S3
  20. Урок 20. 00:13:22
    Exploratory Data Analysis - Part 1
  21. Урок 21. 00:06:08
    Exploratory Data Analysis - Part 2
  22. Урок 22. 00:11:09
    Data Visualization and Best Practices
  23. Урок 23. 00:18:25
    Setting Up Our Training Job Notebook + Reasons to Use SageMaker
  24. Урок 24. 00:13:37
    Python Script for HuggingFace Estimator
  25. Урок 25. 00:03:22
    Creating Our Optional Experiment Notebook - Part 1
  26. Урок 26. 00:04:02
    Creating Our Optional Experiment Notebook - Part 2
  27. Урок 27. 00:13:25
    Encoding Categorical Labels to Numeric Values
  28. Урок 28. 00:15:06
    Understanding the Tokenization Vocabulary
  29. Урок 29. 00:10:57
    Encoding Tokens
  30. Урок 30. 00:12:49
    Practical Example of Tokenization and Encoding
  31. Урок 31. 00:16:57
    Creating Our Dataset Loader Class
  32. Урок 32. 00:15:10
    Setting Pytorch DataLoader
  33. Урок 33. 00:01:32
    Which Path Will You Take?
  34. Урок 34. 00:04:47
    DistilBert vs. Bert Differences
  35. Урок 35. 00:07:41
    Embeddings In A Continuous Vector Space
  36. Урок 36. 00:05:14
    Introduction To Positional Encodings
  37. Урок 37. 00:04:15
    Positional Encodings - Part 1
  38. Урок 38. 00:10:11
    Positional Encodings - Part 2 (Even and Odd Indices)
  39. Урок 39. 00:05:09
    Why Use Sine and Cosine Functions
  40. Урок 40. 00:09:53
    Understanding the Nature of Sine and Cosine Functions
  41. Урок 41. 00:09:25
    Visualizing Positional Encodings in Sine and Cosine Graphs
  42. Урок 42. 00:18:08
    Solving the Equations to Get the Values for Positional Encodings
  43. Урок 43. 00:03:03
    Introduction to Attention Mechanism
  44. Урок 44. 00:18:11
    Query, Key and Value Matrix
  45. Урок 45. 00:06:54
    Getting Started with Our Step by Step Attention Calculation
  46. Урок 46. 00:20:06
    Calculating Key Vectors
  47. Урок 47. 00:10:21
    Query Matrix Introduction
  48. Урок 48. 00:21:25
    Calculating Raw Attention Scores
  49. Урок 49. 00:13:33
    Understanding the Mathematics Behind Dot Products and Vector Alignment
  50. Урок 50. 00:05:43
    Visualizing Raw Attention Scores in 2D
  51. Урок 51. 00:09:17
    Converting Raw Attention Scores to Probability Distributions with Softmax
  52. Урок 52. 00:03:20
    Normalization
  53. Урок 53. 00:09:08
    Understanding the Value Matrix and Value Vector
  54. Урок 54. 00:10:46
    Calculating the Final Context Aware Rich Representation for the Word "River"
  55. Урок 55. 00:01:59
    Understanding the Output
  56. Урок 56. 00:11:56
    Understanding Multi Head Attention
  57. Урок 57. 00:09:52
    Multi Head Attention Example and Subsequent Layers
  58. Урок 58. 00:02:30
    Masked Language Learning
  59. Урок 59. 00:02:57
    Exercise: Imposter Syndrome
  60. Урок 60. 00:17:15
    Creating Our Custom Model Architecture with PyTorch
  61. Урок 61. 00:15:32
    Adding the Dropout, Linear Layer, and ReLU to Our Model
  62. Урок 62. 00:13:05
    Creating Our Accuracy Function
  63. Урок 63. 00:19:09
    Creating Our Train Function
  64. Урок 64. 00:08:18
    Finishing Our Train Function
  65. Урок 65. 00:13:41
    Setting Up the Validation Function
  66. Урок 66. 00:04:06
    Passing Parameters In SageMaker
  67. Урок 67. 00:04:28
    Setting Up Model Parameters For Training
  68. Урок 68. 00:05:40
    Understanding The Mathematics Behind Cross Entropy Loss
  69. Урок 69. 00:06:57
    Finishing Our Script.py File
  70. Урок 70. 00:08:16
    Starting Our Training Job
  71. Урок 71. 00:14:17
    Debugging Our Training Job With AWS CloudWatch
  72. Урок 72. 00:05:47
    Analyzing Our Training Job Results
  73. Урок 73. 00:08:35
    Creating Our Inference Script For Our PyTorch Model
  74. Урок 74. 00:09:13
    Finishing Our PyTorch Inference Script
  75. Урок 75. 00:07:31
    Setting Up Our Deployment
  76. Урок 76. 00:08:55
    Deploying Our Model To A SageMaker Endpoint
  77. Урок 77. 00:04:20
    Introduction to Endpoint Load Testing
  78. Урок 78. 00:10:03
    Creating Our Test Data for Load Testing
  79. Урок 79. 00:01:04
    Upload Testing Data to S3
  80. Урок 80. 00:03:59
    Creating Our Model for Load Testing
  81. Урок 81. 00:07:15
    Starting Our Load Test Job
  82. Урок 82. 00:10:17
    Analyze Load Test Results
  83. Урок 83. 00:03:51
    Deploying Our Endpoint
  84. Урок 84. 00:10:27
    Creating Lambda Function to Call Our Endpoint
  85. Урок 85. 00:05:28
    Setting Up Our AWS API Gateway
  86. Урок 86. 00:05:40
    Testing Our Model with Postman, API Gateway and Lambda
  87. Урок 87. 00:02:52
    Finishing Part 1 and Cleaning Up Resources for Part 1
  88. Урок 88. 00:02:29
    Creating a SageMaker Domain
  89. Урок 89. 00:04:54
    Logging in to our SageMaker Environment
  90. Урок 90. 00:07:38
    Introduction to JupyterLab
  91. Урок 91. 00:07:51
    Sagemaker Sessions, Regions, and IAM Roles
  92. Урок 92. 00:13:30
    Examining Our Dataset from HuggingFace
  93. Урок 93. 00:09:09
    Tokenization and Word Embeddings
  94. Урок 94. 00:04:22
    HuggingFace Authentication with Sagemaker
  95. Урок 95. 00:08:44
    Applying the Templating Function to our Dataset
  96. Урок 96. 00:15:56
    Attention Masks and Padding
  97. Урок 97. 00:04:04
    Star Unpacking with Python
  98. Урок 98. 00:10:23
    Chain Iterator, List Constructor and Attention Mask example with Python
  99. Урок 99. 00:08:12
    Understanding Batching
  100. Урок 100. 00:07:32
    Slicing and Chunking our Dataset
  101. Урок 101. 00:16:07
    Creating our Custom Chunking Function
  102. Урок 102. 00:09:31
    Tokenizing our Dataset
  103. Урок 103. 00:04:31
    Running our Chunking Function
  104. Урок 104. 00:08:33
    Understanding the Entire Chunking Process
  105. Урок 105. 00:05:54
    Uploading the Training Data to AWS S3
  106. Урок 106. 00:06:48
    Setting Up Hyperparameters for the Training Job
  107. Урок 107. 00:06:46
    Creating our HuggingFace Estimator in Sagemaker
  108. Урок 108. 00:08:12
    Introduction to Low-rank adaptation (LoRA)
  109. Урок 109. 00:10:56
    LoRA Numerical Example
  110. Урок 110. 00:09:09
    LoRA Summarization and Cost Saving Calculation
  111. Урок 111. 00:04:46
    (Optional) Matrix Multiplication Refresher
  112. Урок 112. 00:12:33
    Understanding LoRA Programatically Part 1
  113. Урок 113. 00:05:49
    Understanding LoRA Programatically Part 2
  114. Урок 114. 00:08:11
    Bfloat16 vs Float32
  115. Урок 115. 00:06:33
    Comparing Bfloat16 Vs Float32 Programatically
  116. Урок 116. 00:07:20
    Setting up Imports and Libraries for the Train Script
  117. Урок 117. 00:07:57
    Argument Parsing Function Part 1
  118. Урок 118. 00:10:55
    Argument Parsing Function Part 2
  119. Урок 119. 00:14:31
    Understanding Trainable Parameters Caveats
  120. Урок 120. 00:07:36
    Introduction to Quantization
  121. Урок 121. 00:07:20
    Identifying Trainable Layers for LoRA
  122. Урок 122. 00:04:36
    Setting up Parameter Efficient Fine Tuning
  123. Урок 123. 00:10:35
    Implement LoRA Configuration and Mixed Precision Training
  124. Урок 124. 00:04:22
    Understanding Double Quantization
  125. Урок 125. 00:14:15
    Creating the Training Function Part 1
  126. Урок 126. 00:07:17
    Creating the Training Function Part 2
  127. Урок 127. 00:05:09
    Finishing our Sagemaker Script
  128. Урок 128. 00:05:11
    Gaining Access to Powerful GPUs with AWS Quotas
  129. Урок 129. 00:03:55
    Final Fixes Before Training
  130. Урок 130. 00:07:16
    Starting our Training Job
  131. Урок 131. 00:11:24
    Inspecting the Results of our Training Job and Monitoring with Cloudwatch
  132. Урок 132. 00:17:58
    Deploying our LLM to a Sagemaker Endpoint
  133. Урок 133. 00:08:19
    Testing our LLM in Sagemaker Locally
  134. Урок 134. 00:08:56
    Creating the Lambda Function to Invoke our Endpoint
  135. Урок 135. 00:02:37
    Creating API Gateway to Deploy the Model Through the Internet
  136. Урок 136. 00:05:12
    Implementing our Streamlit App
  137. Урок 137. 00:03:27
    Streamlit App Correction
  138. Урок 138. 00:02:39
    Congratulations and Cleaning up AWS Resources
  139. Урок 139. 00:01:18
    Thank You!