[Llama 3.1] Updates dataset, logging, and checkpoint resume. #787
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
MLCommons CLA bot All contributors have signed the MLCommons CLA ✍️ ✅ |
ShriyaRishab
reviewed
Mar 6, 2025
ShriyaRishab
reviewed
Mar 6, 2025
| @@ -103,7 +104,7 @@ After the download is complete, you should see five files under `TOKENIZER_PATH` | |||
|
|
|||
| ### Training and test data separation | |||
|
|
|||
| We use the default split from the C4 dataset. This means that we use `c4-train.<x>-of-01024.json.gz` files for training and `c4-validation.<x>-of-00008.json.gz` files for evaluation. | |||
| We use the default split from the C4 dataset. This means that we use `c4-train.<x>-of-01024.json.gz` files (where `768 <= x <= 1023`) for training, and we use our customized `c4-validation-91205-samples.en.json.gz`, which contains the first 91205 samples from the unshuffled C4 validation dataset, for evaluation. | |||
There was a problem hiding this comment.
Can you explain why and how 91205 samples were chosen?
There was a problem hiding this comment.
This is discussed in the utils/consolidate_data.sh. I can copy the lines here to make it more clear in the README.
There was a problem hiding this comment.
I have added the description in this commit
ShriyaRishab
previously approved these changes
Mar 6, 2025
|
@mlcommons/wg-training @Elnifio Are we good to approve and merge this PR? My team is waiting for this PR to be merged before removing duplicate data from R2. Thanks. |
ShriyaRishab
approved these changes
Mar 18, 2025
|
Thanks! |
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR includes the following changes:
STEP_TIME_ATOLso that, if a training step time is longer than this threshold, then we actively kill this job. Defaults to 2hrs per step.