Commit Graph

552 Commits

Author SHA1 Message Date
Jan Wassenberg f31e12e63b Improved blob diff: parallel, tolerance for float
PiperOrigin-RevId: 724060325
2025-02-06 13:46:28 -08:00
Jan Wassenberg 9f5159ff68 Public visibility for compression/
PiperOrigin-RevId: 723529541
2025-02-05 08:53:51 -08:00
Phil Culliton 7ccc6abe87 Allow conversion, loading and inference with NUQ.
PiperOrigin-RevId: 723507890
2025-02-05 07:45:54 -08:00
Phil Culliton 8a6edff319 Base interleaved handling for 4.5-bit NUQ, specifically Enc, DecompressAndZeroPad, and Dec2. Includes tests.
PiperOrigin-RevId: 721821577
2025-01-31 10:35:32 -08:00
Phil Culliton 23dac72463 Simplified interface class and example for Gemma.cpp usage.
PiperOrigin-RevId: 720591037
2025-01-28 08:48:27 -08:00
Daniel Keysers 7af2e70321 Add python wrappers for configs and inference.
Enable building compression/python/compression_test using bazel.
Add default image path for image_test and paligemma_test.

PiperOrigin-RevId: 720583438
2025-01-28 08:22:03 -08:00
Daniel Keysers bcdb0d65bd Assorted small cleanups.
PiperOrigin-RevId: 720548132
2025-01-28 06:09:45 -08:00
Jan Wassenberg a248f76245 Allow overriding num threads despite detecting topology
PiperOrigin-RevId: 720188756
2025-01-27 08:57:53 -08:00
Daniel Keysers e997468496 Apply PositionalEncodingQK always in-place.
PiperOrigin-RevId: 718851803
2025-01-23 07:09:30 -08:00
Apoorv Reddy ce807a31a1 internal change
PiperOrigin-RevId: 718824952
2025-01-23 05:31:11 -08:00
Jan Wassenberg a60b564b88 Infra improvements (2)
ops.h: move CreateInvTimescale to allow calling without depending on gemma
Pass around MatMulEnv instead of pools to avoid re-creating the env
profiler.h can now be used outside SIMD code
allocator: add StepBytes and QuantumSteps
rename worker thread with package/cluster in the name
threading: add Visit* to IndexRange
PiperOrigin-RevId: 718766704
2025-01-23 01:55:19 -08:00
Daniel Keysers f37402da57 Add parameter for base_frequency to CreateInvTimeScale().
Extract a few local variables to make code easier to read (hopefully).

PiperOrigin-RevId: 718749053
2025-01-23 00:56:44 -08:00
Daniel Keysers a133b3d062 Tiny fix: align template parameter order with parameter order.
PiperOrigin-RevId: 718411494
2025-01-22 09:13:23 -08:00
Phil Culliton 9646edc908 Internal change
PiperOrigin-RevId: 717916568
2025-01-21 07:53:49 -08:00
Copybara-Service f46052b5b4 Merge pull request #473 from ufownl:bugfix/migrate_weights_target
PiperOrigin-RevId: 717536480
2025-01-20 08:05:38 -08:00
Jan Wassenberg c4398fc72d Infra improvements:
allocator: support mmap, fixed Bind, add padding
bench_matmul: Add PreventElision
BUILD: add ops_test build target
matmul.h: move ConstMat here; dynamic alloc of MatMulEnv
matmul_test: remove benchmarking
replace fprintf with HWY_WARN
threading.cc: support splitting large clusters (disabled); package_idx->pkg_idx, smaller IndexRangePartition
PiperOrigin-RevId: 717512274
2025-01-20 06:22:49 -08:00
RangerUFO 20e5ef6d2e Add the missing `migrate_weights` target for CMake 2025-01-17 18:56:43 +08:00
Daniel Keysers 493688f6f1 Allow interactive use with new single-file weight format.
Add section about new weights format to README.md.
Remove model_type_required parameter.
Update error handling for flags.

PiperOrigin-RevId: 715788822
2025-01-15 07:22:33 -08:00
Ray Smith b93231a47d Moved the vit config fields to their own config struct
PiperOrigin-RevId: 715692800
2025-01-15 01:09:49 -08:00
Ray Smith 9d40f0117e Added ability to load/save a complete model file, including tokenizer.
PiperOrigin-RevId: 707914366
2024-12-19 07:59:41 -08:00
The gemma.cpp Authors 5bc356f18f Internal change
PiperOrigin-RevId: 707268913
2024-12-17 15:15:57 -08:00
Daniel Keysers 73766e8ee3 Small updates to the README file.
PiperOrigin-RevId: 707036429
2024-12-17 04:09:55 -08:00
Daniel Keysers 62c70d6715 Rename ModelTraining to PromptWrapping which is a more accurate name.
PiperOrigin-RevId: 705881500
2024-12-13 07:45:59 -08:00
Ray Smith 6254f2e5ca Removed duplicated tensor sizes from weights.h by changing the constructor used for MatPtrT
PiperOrigin-RevId: 705085054
2024-12-11 06:30:28 -08:00
Daniel Keysers aed17396be Make prompt wrapping more consistent and fix duplicated tokens for multi-turn.
Do not echo <end_of_turn> tokens to the user.
Have verbosity=0 only show the dialog.

PiperOrigin-RevId: 705021391
2024-12-11 01:52:00 -08:00
Ray Smith e69bc3bc1c Added the TensorInfo arg to the compressor so the shape and scale can be output correctly to the file in future.
Corrected some errors in the TensorIndex.

PiperOrigin-RevId: 705014619
2024-12-11 01:26:35 -08:00
Jan Wassenberg 7b77909427 Fix unhandled switch warning/error
PiperOrigin-RevId: 704828160
2024-12-10 13:32:53 -08:00
Jan Wassenberg 642fc97d51 Internal change
PiperOrigin-RevId: 704692923
2024-12-10 06:58:32 -08:00
Copybara-Service d8135e836f Merge pull request #460 from ericcurtin:common
PiperOrigin-RevId: 704684454
2024-12-10 06:33:37 -08:00
Daniel Keysers 5bbe814a53 Tiny cleanup.
PiperOrigin-RevId: 704636988
2024-12-10 03:34:05 -08:00
Daniel Keysers 331d2ccc02 Add support for 448px resolution to PaliGemma and PaliGemma2.
PiperOrigin-RevId: 704361579
2024-12-09 11:38:10 -08:00
Eric Curtin a971088ac2 Refactor `gemma/common.cc` to improve readability and safety
Use `std::size` for array size calculations. Replace C-style
string manipulations with `std::string` methods. Simplify
`std::transform` usage for case conversion.

Signed-off-by: Eric Curtin <ecurtin@redhat.com>
2024-12-09 16:36:25 +00:00
The gemma.cpp Authors 66bb435121 No public description
PiperOrigin-RevId: 704178245
2024-12-09 00:49:36 -08:00
Phil Culliton 9dfe2a76be Internal change
PiperOrigin-RevId: 702961613
2024-12-04 20:41:47 -08:00
Jan Wassenberg 6a34e9c547 Print cache info and update Highway version for that
PiperOrigin-RevId: 702318451
2024-12-03 06:31:52 -08:00
Jan Wassenberg f74d496879 Threading/infra improvements.
* Add Parallelize*Range helpers and partitioning helpers
* Refactor Pinning class, store original affinity (required to construct another NestedPools after pinning happened)

Compress:
* prevent Compress printing stats in tests
* zero-pad tensors

Matmul:
* add matmul_unit_test (TODO) and bench_matmul
* matmul_test: change norm to row vectors (that is what is added) and include bf16 rounding error
* Prepare for L2/L3 retrieval
PiperOrigin-RevId: 700603811
2024-11-27 01:12:00 -08:00
Stanko Novakovic 109a4d9f85 Add a simple benchmark for batching.
This is a simple Gemma benchmark with a fixed batch size of 32.

PiperOrigin-RevId: 698843573
2024-11-21 10:59:49 -08:00
Ray Smith 3d1625d8c5 Improved consistency of compressor API, and added a universal method with a target type arg.
Moved configs pybind up to root level.

PiperOrigin-RevId: 698743417
2024-11-21 05:27:40 -08:00
Ray Smith 73640d2521 Added tensor_index as a single source of truth on tensor shapes/sources and transformations
PiperOrigin-RevId: 697903886
2024-11-19 00:25:39 -08:00
Ray Smith 7d685a267f Added pybind for configs.
Added ability to test configs for equality.

PiperOrigin-RevId: 697572671
2024-11-18 04:03:51 -08:00
Jan Wassenberg 36f02ef892 Internal change.
PiperOrigin-RevId: 696815335
2024-11-15 02:22:32 -08:00
Ray Smith 96513a8dc3 Added a blob_compare tool that compares two sbs files that may have the blobs in a different order
PiperOrigin-RevId: 696458888
2024-11-14 03:26:32 -08:00
Paul Chang 5674c33dc5 Replace CLIF SbsWriter with pybind-based gcpp extension
Maintains compatibility with previous version.

PiperOrigin-RevId: 696181603
2024-11-13 10:20:02 -08:00
Daniel Keysers 719699f132 Make top_k a runtime argument (instead of a model argument).
PiperOrigin-RevId: 696170691
2024-11-13 09:48:59 -08:00
Paul Chang b94295b6d9 Internal changes
PiperOrigin-RevId: 696155630
2024-11-13 09:01:38 -08:00
Daniel Keysers e54d9cbddd Fix Griffin model:
- use HalfRope position encodings
- zero-initialize the caches for each Generate at position 0

The lack of the latter made the tests in gemma_test dependent on each other.

PiperOrigin-RevId: 694509054
2024-11-08 08:30:53 -08:00
Paul Chang d4050a2917 Expose BlobReader::Keys()
PiperOrigin-RevId: 694166186
2024-11-07 10:28:39 -08:00
Jan Wassenberg 868b01601f Simpler MatMul interface, vocab types, Tristate for use_spinning
Add Extents2D, Range2D vocab types
Matmul uses ConstMat for inputs and RowPtr for output
Move RowVectorBatch to basics.h
Separate threading.cc
Fix topology string: report cores not LPs, and #HT
Move QStride/IsMHA into LayerConfig
ImageTokens does not require make_unique.
matmul_test: no longer require template args
PiperOrigin-RevId: 692963605
2024-11-04 07:48:29 -08:00
Jan Wassenberg baaa221787 Move BF16 to basics.h for easier access, and use that typedef.
PiperOrigin-RevId: 691422334
2024-10-30 08:09:11 -07:00
Daniel Keysers ed4091921f Reduce time for optimize_test and use exactly one (unpinned) thread.
PiperOrigin-RevId: 691013413
2024-10-29 07:37:22 -07:00