google / gemma.cpp

lightweight, standalone C++ inference engine for Google's Gemma models.

google/gemma.cpp Issues

How to make gemma support Chinese?
Closed 3 months ago7
Near-term roadmap
Updated 7 days ago3
Have there been any performance comparisons between gemma.cpp and llama.cpp?
Updated 8 days ago3
Paligemma Support
Updated 8 days ago1
Wrong keyboard mapping in command line
Closed 17 days ago3
How can I get the compiled static executable binary file in linux platform?
Closed 17 days ago2
Use a MatMul implementation over MatVec for Prefill Computations
Updated 21 days ago2
Mismatch between expected 1024000 and actual 1024512 KiB size.
Updated 24 days ago1
AVX2 is dimwitted compared to AVX512
Updated 25 days ago8
Android aarch64 support
Closed 3 months ago8
the 7b-pt-sfp doesn‘t looks like normal. some thing wrong with the args?
Closed 3 months ago2
WSL support: "Killed", once executed on WSL
Closed 3 months ago9
num_threads doesn't seem to have any effect
Closed 3 months ago6
Incorporating gemma.cpp as a Library in Android Project
Closed 3 months ago7
Failed to read from model.weights.h5 - might be a directory, or too small?
Closed 3 months ago3
compile problems
Closed 3 months ago6
Compilation problem on WSL2
Closed 3 months ago1
How to use bfloat16_t to build parameters in Windows. Using the 7b-it model
Closed 2 months ago1
The continuity of the conversation is problematic.
Closed 3 months ago2
How to open PROFILER_ZONE to see the hwy profiling results?
Closed 3 months ago2
Running make -j4 gemma fails on android termux
Closed 3 months ago2
Error during build process
Closed 2 months ago1
Compile error on visual studio build, array size exceed
Closed 2 months ago2
gemma.exe does not respond no matter how long you wait.
Closed 2 months ago3
Using mingw64 build on windows 10 fails
Closed 2 months ago3
Is the format of output Markdown?
Closed a month ago2
again:Cached compressed weights does not exist yet gemma
Updated a month ago5
make error on RaspberryPi 4
Closed 3 months ago3
request to remove `cmake`, `highway`, and `gtest`
Closed a month ago10
The application does not use Nvidia GPU in WSL
Closed 3 months ago2
make error on orangepi 5 (arm)
Closed 3 months ago6
To do
Closed 3 months ago1
Error at the end when compiling
Closed 3 months ago7
Functions exposed in the libgemma.a?
Closed 3 months ago2
Building android executables on ubuntu
Closed 3 months ago2
project does not work as submodule
Closed 2 months ago2
Crash in Attention() with huge input
Closed 2 months ago13
Weights converted from PyTorch don't seem to work properly
Closed 2 months ago2
Implement arch changes
Closed a month ago
Failed on raspberry pi OS (64bit)
Closed a month ago6
TODO (**Optimize, potentially using new VQSort PartialSort**)
Updated a month ago1
Any more fine example or user-guide for CodeGemma?
Updated a month ago
Seeking feedbacks for python wrapper of gemma.cpp
Closed 2 months ago3
How to use weights on HuggingingFace?
Closed 3 months ago4
Add Self-Extend to the gemma.cpp
Updated 3 months ago4
Gemma.cpp on the Android arm64-v8a
Closed 3 months ago4
the --log option doesn't actually save the output to a log file
Closed 3 months ago2
GRPC support - in scope?
Updated 3 months ago2
Make Copybara close pull requests upon merge, use more informative commit descriptions
Closed 3 months ago1
[Suggestions] Low effort OpenMP, OpenACC, CBLAS compatible CPU & GPU acceleration + other improvements
Updated 3 months ago1