Support GGUF Mixtral format where experts are in one tensor #355
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
Code Metrics Report=============================================================================== Language Files Lines Code Comments Blanks =============================================================================== Dockerfile 1 34 25 0 9 Happy 1 442 369 0 73 JSON 9 21 21 0 0 Python 21 741 622 21 98 TOML 15 393 355 1 37 ------------------------------------------------------------------------------- Jupyter Notebooks 1 0 0 0 0 |- Markdown 1 60 30 22 8 |- Python 1 96 87 1 8 (Total) 156 117 23 16 ------------------------------------------------------------------------------- Markdown 16 1054 0 781 273 |- BASH 6 203 190 0 13 |- Python 6 121 110 0 11 |- Rust 3 185 172 9 4 (Total) 1563 472 790 301 ------------------------------------------------------------------------------- Rust 86 28434 26022 377 2035 |- Markdown 42 439 0 427 12 (Total) 28873 26022 804 2047 =============================================================================== Total 151 31119 27414 1180 2525 =============================================================================== |
EricLBuehler
added a commit
that referenced
this pull request
Jun 10, 2024
EricLBuehler
added a commit
that referenced
this pull request
Jun 10, 2024
* Implement gpt2 gguf tokenizer * Fix unk tok calculation * Remove normalizer * Update gguf tokenizer * Allow adding unk token when found * Add unk token to builder if provided. * Improve add_special_tokens * Use tokenizerx builder * Add useful comment Co-authored-by: Brennan Kinney <5098581+polarathene@users.noreply.github.com> * Bump version to 0.1.16 (#404) * Bump version to 0.1.17 * Fix version bump * Add and update template READMEs (#405) * Add readmes * Fix typos * Improve Rust docs (#406) * Expose phi3v loader and remove unused deps (#408) * Support format for mixtral where experts are in one tensor (#355) * Normal loading metadata for vision models (#409) * Phi 3 vision ISQ support (#410) * ISQ support for phi3v * Document it * Remove causal masks cache (#412) * Fix: use new slice_assign (#415) * Use new slice_assign * Fix dead image links * Fix Phi-3 GGUF (#414) * Fix kv head usage * Fix rope weights * Clippy * Work on the gpt2 conversion * Add comment * Add some tests * Update readme --------- Co-authored-by: Brennan Kinney <5098581+polarathene@users.noreply.github.com>
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.
Refs #352. After this PR, both the
ffn_*_expsandffn_*.{i}formats for expert tensors will be accepted.