Repository navigation
Move the semantic search out of the open core; embed the corpus with Gradle - #95
Merged
Merged
Conversation
The embedding model, its embedder and the hybrid engine now live in the private seforim-semantic package (Zayit's official builds). The public bundle no longer carries the model, and DownloadEmbedModel is gone. The generator still indexes dense vectors given -PvectorsBin. Signed-off-by: Elie Gambache <elyahou.hadass@gmail.com>
…tic package is readable With the token that reads seforim-semantic (the siddur's, as in Zayit, or gh), buildLuceneIndexDefault depends on embedCorpus, which runs the package's EmbedCorpus: fp32 model on an NVIDIA GPU with the CUDA 12 / cuDNN 9 runtime fetched from NVIDIA's redistributables (-PcudaArchivesDir for archives at hand), else int8 on the CPU. No Python. Community builds stay lexical only. Signed-off-by: Elie Gambache <elyahou.hadass@gmail.com>
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
seforim-semanticpackage (Zayit's official builds). The public DB bundle no longer carries the model;DownloadEmbedModelis removed.:searchindex:embedCorpus: with the token that readsseforim-semantic(the siddur's, orgh),buildLuceneIndexDefaultembeds the corpus into the index's dense vectors, with no Python. fp32 on an NVIDIA GPU (Linux x64), with CUDA 12.8 / cuDNN 9.10 fetched from NVIDIA's redistributables (sha256-checked, cached in the Gradle user home,-PcudaArchivesDirfor archives at hand), else int8 on the CPU.-PvectorsBinand-PnoVectorsstill apply.Test plan
embedCorpuson a 13,372-line sample: GPU, ~900 lines/s; vectors identical to the former Python ones (cosine mean 1.00000, min 0.99999), same books and base flags:searchcompiles and tests pass