multi-threaded model initialization #737
Open
Conversation
This file contains 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
ademeure
reviewed
Aug 12, 2024
| @@ -504,6 +504,47 @@ void gpt2_build_from_checkpoint(GPT2 *model, const char* checkpoint_path, bool w | |||
| cudaCheck(cudaDeviceSynchronize()); | |||
| } | |||
|
|
|||
| void gpt2_init_layer(GPT2 *model, int l, mt19937_state* rng, floatX* params) { | |||
| int offset = 0; | |||
There was a problem hiding this comment.
This should be size_t for larger model sizes.
|
Looks good to me, it improves startup time for -e "d72" from ~100s to ~15s on a 1xH100 node with 26 CPU cores! :) |
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Outdated suggestions cannot be applied.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.
A simple demonstration of initializing from multiple threads.
Factored out the parallel part; by making it a separate function, it becomes less likely that we accidentally change shared state by using variables from the outer scope.