January 2025 Update #1036
Merged
January 2025 Update #1036
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
…eeds more changes for other backends
|
Thanks for your hard work. Any chance this can be updated with the latest binaries? Mainstream got QRWKV Hybrid, Phi-3.5-MoE, Deepseek V3, C4AI Command-R 7B models, INT-8 and BF16 implementations. |
|
The binaries here are from just one week ago, so hopefully most of those things should be covered. Currently we need to fix the issues loading the binaries on Linux and MacOS, once that's figured out the next update should be easier 🤞 |
Backend library split compatibility for January 2025 binary update
Wip december update fixes v2
…er defined AVX level
Wip december update fixes v3
|
Unit tests passed on Windows CUDA & Linux CUDA. Test application is running fine on:
|
|
Works on my machine! (osx-arm64)
Tested all of |
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Outdated suggestions cannot be applied.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.
Updated llama.cpp binaries to 0827b2c1da299805288abbd556d869318f2b121e.
This introduces new binaries. Previously there were 2 for each platform, e.g.
ggml.dll, llama.dllfor CUDA. Now there are more:ggml.dll, ggml-base.dll, ggml-cpu.dll, ggml-cuda.dll, llama.dll.Currently these are handled in the same way as the old system - each platform has it's own set of completely independent binaries. In the future this should be modified to dynamically load backends. See more details here and here.
Testing: