Add MaxText Llama 3.1 70B training with GCS recipe #56
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
raymondzouu
force-pushed
the
storage-recipe
branch
from
698777c to
57ff296
Compare
mkmg
suggested changes
Apr 9, 2025
training/trillium/Llama3.1-70B-MaxText-with-Storage/checkpoint_pvc.yaml
Outdated
Show resolved
Hide resolved
training/trillium/Llama3.1-70B-MaxText-with-Storage/dataset_pvc.yaml
Outdated
Show resolved
Hide resolved
training/trillium/Llama3.1-70B-MaxText-with-Storage/dataset_pvc.yaml
Outdated
Show resolved
Hide resolved
training/trillium/Llama3.1-70B-MaxText-with-Storage/dataset_pvc.yaml
Outdated
Show resolved
Hide resolved
...ning/trillium/Llama3.1-70B-MaxText-with-Storage/llama3-1-70B-real-data-and-ckpt-1xv6e-256.sh
Outdated
Show resolved
Hide resolved
akansha1812
reviewed
Apr 9, 2025
akansha1812
reviewed
Apr 9, 2025
raymondzouu
force-pushed
the
storage-recipe
branch
4 times, most recently
from
f8dcddc to
51d9cee
Compare
akansha1812
reviewed
Apr 10, 2025
| @@ -0,0 +1,11 @@ | |||
| python3 benchmarks/benchmark_runner.py xpk \ | |||
| project=$PROJECT \ | |||
| zone=$ZONE \ | |||
There was a problem hiding this comment.
Not sure why we need the .sh script here. This is not blocking, if you think its needed, go ahead with it.
LGTM.
akansha1812
reviewed
Apr 10, 2025
akansha1812
reviewed
Apr 10, 2025
akansha1812
reviewed
Apr 10, 2025
training/trillium/Llama3.1-70B-MaxText-with-Storage/dataset_pvc.yaml
Outdated
Show resolved
Hide resolved
akansha1812
reviewed
Apr 10, 2025
training/trillium/Llama3.1-70B-MaxText-with-Storage/checkpoint_pvc.yaml
Outdated
Show resolved
Hide resolved
raymondzouu
force-pushed
the
storage-recipe
branch
2 times, most recently
from
2246636 to
c568eb0
Compare
raymondzouu
force-pushed
the
storage-recipe
branch
from
c568eb0 to
fe2b53f
Compare
mkmg
suggested changes
Apr 10, 2025
raymondzouu
force-pushed
the
storage-recipe
branch
2 times, most recently
from
f9d5280 to
9f57b66
Compare
mkmg
approved these changes
Apr 10, 2025
akansha1812
approved these changes
Apr 10, 2025
raymondzouu
force-pushed
the
storage-recipe
branch
from
9f57b66 to
e52b1b8
Compare
raymondzouu
force-pushed
the
storage-recipe
branch
from
e52b1b8 to
4aea327
Compare
mkmg
reviewed
May 15, 2025
| device_type=v6e-256 \ | ||
| num_slices=1 \ | ||
| cluster_name=$CLUSTER \ | ||
| base_output_directory=$OUTPUT_DIR \ |
There was a problem hiding this comment.
I just noticed that there are some env variables here that users are not instructed to populate above.
mkmg
reviewed
May 15, 2025
| ``` | ||
| export RECIPE_REPO="path-to-this-recipe-repo" # Update | ||
|
|
||
| cd ~/xpk |
There was a problem hiding this comment.
Also we never instruct the user to clone XPK.
There was a problem hiding this comment.
In step 1 we link to the instructions for cloning xpk and installing the dependencies.
raymondzouu
force-pushed
the
storage-recipe
branch
from
4aea327 to
2e13786
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
No description provided.