github-starred/cs249r_book
2
0
Fork 0
mirror of https://github.com/harvard-edge/cs249r_book.git synced 2026-07-17 16:34:48 -05:00
Code Issues 276 Packages Projects Releases 26 Wiki Activity

[GH-ISSUE #1685] [TinyTorch] Concept check in Module 1 docs is checking knowledge on something not covered #13545

New Issue
Closed
opened 2026-05-17 17:42:02 -05:00 by GiteaMirror · 5 comments
GiteaMirror commented 2026-05-17 17:42:02 -05:00
Owner
Copy Link

Originally created by @bdub-1 on GitHub (May 5, 2026).
Original GitHub issue: https://github.com/harvard-edge/cs249r_book/issues/1685

Module

01 Tensor

Type of Improvement

Better explanation or documentation

Description

https://mlsysbook.ai/tinytorch/modules/01_tensor.html#check-your-understanding:
"How a scalar logit of 32 × 3 × 224 × 224 image batch lands at ~18.4 MB in float32, and why batch size is the first knob to turn when OOM hits."

Proposed Solution

Add documentation on OOM issues

Originally created by @bdub-1 on GitHub (May 5, 2026). Original GitHub issue: https://github.com/harvard-edge/cs249r_book/issues/1685 ### Module 01 Tensor ### Type of Improvement Better explanation or documentation ### Description https://mlsysbook.ai/tinytorch/modules/01_tensor.html#check-your-understanding: "How a scalar logit of 32 × 3 × 224 × 224 image batch lands at ~18.4 MB in float32, and why batch size is the first knob to turn when OOM hits." ### Proposed Solution Add documentation on OOM issues
GiteaMirror added the area: tinytorchtype: improvement labels 2026-05-17 17:42:02 -05:00
GiteaMirror commented 2026-05-17 17:42:04 -05:00
Author
Owner
Copy Link

@Shashank-Tripathi-07 commented on GitHub (May 5, 2026):

Thanks for raising this! The 18.4 MB / OOM explanation is actually already present in the module, it's in the collapsible Q&A section directly below the checklist.

On the published page (01_tensor.html#check-your-understanding), click the "Answer" toggle under Q1: Memory Calculation to expand it:

32 × 3 × 224 × 224 × 4 = 19,267,584 bytes ≈ 18.4 MB

Double the batch, double the memory, which is why batch size is the first knob
people turn when training runs out of GPU memory.

The full derivation is there, it's just collapsed by default so students try to work it out from the checklist first, then verify. No missing content. Closing unless you feel the UX of the collapse is confusing enough to warrant surfacing it differently.

<!-- gh-comment-id:4380943975 --> @Shashank-Tripathi-07 commented on GitHub (May 5, 2026): Thanks for raising this! The 18.4 MB / OOM explanation is actually already present in the module, it's in the collapsible Q&A section directly below the checklist. On the published page ([01_tensor.html#check-your-understanding](https://mlsysbook.ai/tinytorch/modules/01_tensor.html#check-your-understanding)), click the **"Answer"** toggle under **Q1: Memory Calculation** to expand it: ``` 32 × 3 × 224 × 224 × 4 = 19,267,584 bytes ≈ 18.4 MB Double the batch, double the memory, which is why batch size is the first knob people turn when training runs out of GPU memory. ``` The full derivation is there, it's just collapsed by default so students try to work it out from the checklist first, then verify. No missing content. Closing unless you feel the UX of the collapse is confusing enough to warrant surfacing it differently.
GiteaMirror commented 2026-05-17 17:42:05 -05:00
Author
Owner
Copy Link

@bdub-1 commented on GitHub (May 5, 2026):

I was more or less thinking that it was a new concept (I had to look up what OOM stood for). I naturally would search for an issue like this within the "common-errors-debugging" section. That's why I thought it was missing. The phrase "and why batch size is the first knob..." makes me feel like I was suppose to have read why earlier. I did not know that batch size was the first thing to change until reading that.

<!-- gh-comment-id:4381070386 --> @bdub-1 commented on GitHub (May 5, 2026): I was more or less thinking that it was a new concept (I had to look up what OOM stood for). I naturally would search for an issue like this within the "common-errors-debugging" section. That's why I thought it was missing. The phrase "and why batch size is the first knob..." makes me feel like I was suppose to have read why earlier. I did not know that batch size was the first thing to change until reading that.
GiteaMirror commented 2026-05-17 17:42:05 -05:00
Author
Owner
Copy Link

@Shashank-Tripathi-07 commented on GitHub (May 6, 2026):

Valid point. The checklist item used "OOM" without ever defining it, so a student seeing it for the first time had no anchor for the term. The Q&A answer below does explain "runs out of GPU memory" but a student reading the checklist first would not have that context yet.

Fixed in PR #1687: the checklist item now reads "when the GPU runs Out of Memory (OOM)" so the acronym is defined inline and the item is fully self-contained.

<!-- gh-comment-id:4384853426 --> @Shashank-Tripathi-07 commented on GitHub (May 6, 2026): Valid point. The checklist item used "OOM" without ever defining it, so a student seeing it for the first time had no anchor for the term. The Q&A answer below does explain "runs out of GPU memory" but a student reading the checklist first would not have that context yet. Fixed in PR #1687: the checklist item now reads "when the GPU runs Out of Memory (OOM)" so the acronym is defined inline and the item is fully self-contained.
GiteaMirror commented 2026-05-17 17:42:06 -05:00
Author
Owner
Copy Link

@Shashank-Tripathi-07 commented on GitHub (May 6, 2026):

Cool, I've made the changes in a new PR to reduce the drag for learners !

We need the learners to build confidence through the projects and not feel behind in anything caused by our work. Thanks for the review. You just helped us a lot !!

<!-- gh-comment-id:4384860331 --> @Shashank-Tripathi-07 commented on GitHub (May 6, 2026): Cool, I've made the changes in a new PR to reduce the drag for learners ! We need the learners to build confidence through the projects and not feel behind in anything caused by our work. Thanks for the review. You just helped us a lot !!
GiteaMirror commented 2026-05-17 17:42:07 -05:00
Author
Owner
Copy Link

@bdub-1 commented on GitHub (May 6, 2026):

No problem!

<!-- gh-comment-id:4385015851 --> @bdub-1 commented on GitHub (May 6, 2026): No problem!
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
area: book
area: collabs
area: kits
area: labs
area: socratiq
area: staffml
area: tinytorch
area: tools
area: website
bug
dependencies
format: epub
format: pdf
javascript
link-health
link-rot
priority-high
pull-request
Mirrored from GitHub Pull Request
 staffml
type: bug
type: citation
type: code
type: errata
type: improvement
type: new
type: question
vault-sli
No Label area: tinytorch type: improvement
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/cs249r_book#13545
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?