Documented CUDA reproducibility, added warning #1346
Merged
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Outdated suggestions cannot be applied.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.
Fixes #1340 .
When I worked on CUDA acceleration I noticed that the text I generated was not the same as master despite setting the same seed. It seems that due to the use of multiple CUDA streams, the reproducibility of cuBLAS matrix multiplication is not guaranteed . This does not seem to affect perplexity scores but it can unnecessarily cost developers time when they search for non-existent bugs.
This PR documents the non-reproducible nature of cuBLAS. It adds warnings to the README and
ggml-cuda.cu
. Additionally, when using cuBLAS and setting a seed a warning is printed which states that reproducibility is not guaranteed.