See https://arxiv.org/abs/2407.07726 for a description of the model.
Because PaliGemma operates as a prefix-LM on the image+prompt, add support for that.
PiperOrigin-RevId: 677841119
Compression:
* Implement {any packed} x {bf16, f32} 'Load2' and DecompressAndZeroPad
* New compression test for all packed formats, add to GEMMA_TEST_FILES, remove from sfp/nuq_test
* Decompress->DecompressAndZeroPad, use PackedSpan for args with bounds checking
* NUQ: support arbitrary-length enc/dec
* New compression/shared, remove sfp.h and nuq.h
* Move Store2 into Traits and provide Compress2 wrapper
* Remove unused Decompress()-with-pool overload
* Simplify CompressedArrayLen, rename to CompressedArrayElements
* Remove unused DistortionStats b_l1_
Misc:
* Add compensated and Kahan dot, support any length
* Use same Dot function everywhere
* Move exact arithmetic functions into fp_arith
* use FloatPtr and MatPtr typedefs in tests; less stack usage
* Rename args to packed/raw
* Remove Traits::Name, instead TypeName<T>()
* Move kMaxSFP and kClusters/kGroupSize into Sfp/NuqStream
PiperOrigin-RevId: 672868468
Use args only in GemmaEnv constructor, store everything else in RuntimeConfig.
Add runtime option to turn off thread spinning.
PiperOrigin-RevId: 670467320
Supports converting all weight/activation formats to native MulT (bf16/f32)
Also:
- ConstMat/MutableMat for const correctness
- Move RowVectorBatch to allocator.h so it can be used from Matmul
- Add matmul.h so MatMulEnv can be used from Activations
- Remove kMaxThreads, detect from PerClusterPools
- Build fix: -inl.h files must be textual_hdrs, and highway.h should precede -inl.h
```
zen4 new
64, 24576, 3072, add=0, MatTA=bf16, MatTB=sfp: 616.6 GFLOPS.
64, 3072, 24576, add=0, MatTA=bf16, MatTB=sfp: 460.7 GFLOPS.
64, 24576, 3072, add=0, MatTA=f32, MatTB=sfp: 598.6 GFLOPS.
64, 3072, 24576, add=0, MatTA=f32, MatTB=sfp: 435.6 GFLOPS.
zen4 old
64, 24576, 3072, add=0, MatTA=f32, MatTB=sfp: 257.5 GFLOPS.
64, 3072, 24576, add=0, MatTA=f32, MatTB=sfp: 231.9 GFLOPS.
```
PiperOrigin-RevId: 663729812
Limit thread counts to detected. Add max_clusters arg.
Update detection logic to check for smt0 - previously we pinned to some siblings.
PiperOrigin-RevId: 659755311
- Allocate only the required KV caches and activation batch size
- Add flags for batch sizes
- Const-correct interface: Span of const int.
- Also clean up the KVCache arg to a span.
- Move kPrefillBatchSize into RuntimeConfig and remove related global constants.
PiperOrigin-RevId: 655893197
Helper functions to tokenize/wrap
Move LayersOutputFunc into RuntimeConfig
AcceptFunc passes the probability
Implement StringFromType using the parser, and verify results match
PiperOrigin-RevId: 643255119
accept_token: allow default, check if empty when using
allow mixing sample_func and stream_func, call the latter after the former
Also fix missing includes/deps.
PiperOrigin-RevId: 642240012
This changes the command line flags, but the default value retains the previous behavior.
Also add a CreateGemma helper to enable extra args without interface changes.
PiperOrigin-RevId: 641266411
Move Path into io.h and use for opening files.
Removes dependency of gemma_lib on args.
Separate Windows codepath instead of emulating POSIX functions.
Plus lint fixes.
PiperOrigin-RevId: 626279004
Also implement support for some model variations:
- Local attention.
- Add support for biases.
- Use RoPE only on half vectors.
- Support different order of QKV weights.
Co-authored-by: Andrey Mikhaylov <amik@google.com>
Co-authored-by: Martin Bruse <zondolfin@gmail.com>
Co-authored-by: Zoltan Szabadka <szabadka@google.com>
Also add a script to help running sanitizer builds, and do some cleanup.
Co-authored-by: Andrey Mikhaylov <amik@google.com>
Co-authored-by: Eugene Kliuchnikov <eustas@google.com>
Co-authored-by: Sami Boukortt <sboukortt@google.com>
Co-authored-by: Zoltan Szabadka <szabadka@google.com>
- Allow scaling of SFP weights
- Allow using uncompressed weights
- Do not try to compress weights in the main model calls
- Reduce code duplication in weight handling with some macros
Co-authored-by: Eugene Kliuchnikov <eustas@google.com>
Co-authored-by: Thomas Fischbacher <tfish@google.com>
Co-authored-by: Zoltan Szabadka <szabadka@google.com>
When the user runs "--help", they see the non-existent word
"compressingnew". This is because of a missing space, which
is now added, resulting in "compressing new".
--
19694e1f2e by Silvio Traversaro <silvio@traversaro.it>:
Do not pass explicitly -O2 flag to compiler in Release build
COPYBARA_INTEGRATE_REVIEW=https://github.com/google/gemma.cpp/pull/3 from traversaro:patch-1 19694e1f2e
PiperOrigin-RevId: 610096914
It's not possible to build `gemma.cpp` with the standard MSVC front-end
as it doesn't support arrays more than `0x7ffffffff` bytes (see Compiler Error C2148),
however this isn't a problem with the optional Visual Studio Clang/LLVM frontend.
This can be specified using the `-T` flag when running CMake:
```
$ cmake -B build -T ClangCL
$ cmake --build build --config Release
```
Windows doesn't provide `pread`/`pwrite` so this must be emulated using
the `ReadFile`/`WriteFile` Win32 APIs.
`_CRT_SECURE_NO_WARNINGS` is defined to prevent a large number of warnings
about using "depricated" function names (e.g. `close` instead of `_close`).
`NOMINMAX` is defined to prevent the `min`/`max` macros from `windows.h`
from conflicting with expressions like `std::min`. Generally libraries should
avoid including `windows.h` in their public headers or define `WIN32_LEAN_AND_MEAN`
before including the `windows.h` header, but this unfortunately isn't always the case.
Daniel Keysers
RangerUFO
Jan Wassenberg
The gemma.cpp Authors
Zelalem Aweke
Phil Culliton
Luca Versari
Zoltan Szabadka
Copybara-Service
Eric Ye
Ville Sundell
pculliton
austinvhuang
Dan Zheng
Silvio Traversaro
Yuta Hayashibe
David Coles
Austin Huang