]>
git.djapps.eu Git - pkg/ggml/sources/ggml/log
Georgi Gerganov [Sat, 12 Jul 2025 11:39:52 +0000 (14:39 +0300)]
sync : resolve conflicts (#0)
ggml-ci
Georgi Gerganov [Sat, 12 Jul 2025 11:36:32 +0000 (14:36 +0300)]
sync : llama.cpp
ggml-ci
Jeff Bolz [Sat, 12 Jul 2025 10:12:26 +0000 (05:12 -0500)]
vulkan: support SET_ROWS (llama/14587)
* vulkan: support SET_ROWS
Add variants of the copy_to_quant shader that do the SET_ROWS operation.
Change these shaders to spread the work across the workgroup.
The memory access pattern is probably not great (one thread per quant block),
but should be fine for now.
* vulkan: optimize set_rows
Larger workgroups for non-quant types.
Set "norepeat" (there is manual repeat logic).
Use fastmod.
Jeff Bolz [Sat, 12 Jul 2025 09:51:58 +0000 (04:51 -0500)]
vulkan: optimizations for deepseek prompt processing (llama/14555)
* vulkan: allow unclamped loads in coopmat2 mul_mat_id shader
* vulkan: increase coopmat2 mul_mat_id tile size
* vulkan: optimize mat_mul_id row_ids search to batch loads, and port to coopmat1 path
* vulkan: use smaller FA row size when head size is large. applies to both scalar and CM2 paths (CM1 isn't used due to shared memory limits)
Tarek Dakhran [Fri, 11 Jul 2025 18:27:01 +0000 (20:27 +0200)]
model : support LiquidAI LFM2 hybrid family (llama/14620)
**Important**
LFM2 was [merged ](https://github.com/huggingface/transformers/pull/39340)into transformers, but has not yet been released.
To convert into gguf, install transformers from source
```shell
pip install "transformers @ git+https://github.com/huggingface/transformers.git@main"
```
Slobodan Josic [Fri, 11 Jul 2025 16:55:00 +0000 (18:55 +0200)]
HIP : Add HIP 7.0+ compatibility for hipBLAS compute types (llama/14634)
rmatif [Thu, 10 Jul 2025 21:58:12 +0000 (23:58 +0200)]
opencl: add tiled mul_mat_f16_f32 (llama/14535)
* add tiled mul_mat_f16_f32
* fix trailing whitespace
* add insightful comments
lhez [Thu, 10 Jul 2025 18:48:52 +0000 (11:48 -0700)]
opencl: add `set_rows` for `f16` and `f32` (llama/14547)
* opencl: add `set_rows` for `f16` and `f32`
* opencl: better choose workgroup size for `set_rows`
Aman Gupta [Thu, 10 Jul 2025 15:29:01 +0000 (23:29 +0800)]
Docs: script to auto-generate ggml operations docs (llama/14598)
* Docs: script to auto-generate ggml operations docs
* Review: formatting changes + change github action
* Use built-in types instead of typing
* docs : add BLAS and Metal ops
---------
Co-authored-by: Georgi Gerganov <redacted>
Akarshan Biswas [Thu, 10 Jul 2025 08:29:38 +0000 (13:59 +0530)]
SYCL: Initial set_rows kernel implementation (llama/14562)
* SYCL: Initial set_rows kernel implementation
* Revert max_threads to 256
* Refactor set_rows and address review comments
* Deduplicate conversion function
* Remove guard before kernel launch and refactor
* Fix and add back SFINAE
compilade [Thu, 10 Jul 2025 03:54:38 +0000 (23:54 -0400)]
cuda : support Falcon-H1 state size for SSM_SCAN (llama/14602)
Xuan-Son Nguyen [Wed, 9 Jul 2025 16:16:12 +0000 (18:16 +0200)]
ggml : add ggml_scale_bias (llama/14417)
* ggml : add ggml_scale_bias
* ggml_vec_mad1_f32
* add more simd
* add CUDA
* sycl
* vulkan
* cann (placeholder)
* opencl
* will this fix cpu?
* fix cuda
* suggestions from coderabbit
* fix cann compile error
* vDSP_vsmsa
* rm __ARM_FEATURE_SVE
* use memcpy for op params
* make code looks more consistent
* use scalar for __ARM_FEATURE_SVE
* add x param to ggml_vec_mad1_f32
Miaoqian Lin [Wed, 9 Jul 2025 12:33:53 +0000 (20:33 +0800)]
ggml : prevent integer overflow in gguf tensor size calculation (llama/14595)
Jeff Bolz [Tue, 8 Jul 2025 18:11:42 +0000 (13:11 -0500)]
vulkan: optimize flash attention split_k_reduce (llama/14554)
* vulkan: allow FA split_k with smaller KV values
* vulkan: spread split_k_reduce work across more threads
k_num can get rather large. Use the whole workgroup to reduce the M/L values.
Launch a thread for each element in the HSV dimension of the output. Helps a
lot for large HSV (like deepseek).
Jeff Bolz [Tue, 8 Jul 2025 13:21:21 +0000 (08:21 -0500)]
vulkan : fix rope with partial rotation and non-cont src (llama/14582)
Georgi Gerganov [Tue, 8 Jul 2025 07:15:21 +0000 (10:15 +0300)]
cuda : fix rope with partial rotation and non-cont src (llama/14580)
* cuda : fix rope non-cont
ggml-ci
* cont : fix multi-rope + add test
ggml-ci
* sycl : try fix
ggml-ci
* cont : fix sycl + clean-up cuda
ggml-ci
Aman Gupta [Tue, 8 Jul 2025 02:11:18 +0000 (10:11 +0800)]
CUDA: add bilinear interpolation for upscale (llama/14563)
R0CKSTAR [Mon, 7 Jul 2025 23:58:30 +0000 (07:58 +0800)]
musa: fix build warnings (unused variable) (llama/14561)
Signed-off-by: Xiaodong Ye <redacted>
Aman Gupta [Mon, 7 Jul 2025 13:45:43 +0000 (21:45 +0800)]
CUDA: add bf16 and i32 to getrows (llama/14529)
Eve [Sun, 6 Jul 2025 10:29:36 +0000 (10:29 +0000)]
vulkan: increase LOAD_VEC_A to 8 (IQ1/IQ2) or 4 (IQ3) (llama/14485)
Commit taken from remyoudompheng's PR https://github.com/ggml-org/llama.cpp/pull/12260
Co-authored-by: Rémy Oudompheng <redacted>
Jeff Bolz [Sun, 6 Jul 2025 08:08:16 +0000 (03:08 -0500)]
vulkan: fix rms_norm+mul fusion (llama/14545)
The fused operation was grabbing the epsilon value from the wrong place.
Add an env var to disable fusion.
Add some missing checks for supported shapes/types.
Handle fused rms_norm+mul in check_results.
Jeff Bolz [Sat, 5 Jul 2025 07:26:04 +0000 (02:26 -0500)]
vulkan: Handle updated FA dim2/3 definition (llama/14518)
* vulkan: Handle updated FA dim2/3 definition
Pack mask boolean and n_head_log2 into a single dword to keep the push
constant block under the 128B limit.
* handle null mask for gqa
* allow gqa with dim3>1
Sigbjørn Skjæret [Sat, 5 Jul 2025 06:24:56 +0000 (08:24 +0200)]
opencl: add GELU_ERF (llama/14476)
R0CKSTAR [Sat, 5 Jul 2025 04:10:53 +0000 (12:10 +0800)]
test-backend-ops: add support for specifying output format (llama/14368)
* test-backend-ops: add support for specifying output format
Signed-off-by: Xiaodong Ye <redacted>
* Address review comments
Signed-off-by: Xiaodong Ye <redacted>
* Add build_commit and build_number in test_result
Signed-off-by: Xiaodong Ye <redacted>
* Address review comments
Signed-off-by: Xiaodong Ye <redacted>
* refactor
Signed-off-by: Xiaodong Ye <redacted>
* Get build commit from ggml_commit()
Signed-off-by: Xiaodong Ye <redacted>
* Merge errors into test_operation_info && address review comments
Signed-off-by: Xiaodong Ye <redacted>
* Address review comments
Signed-off-by: Xiaodong Ye <redacted>
* Address review comments
Signed-off-by: Xiaodong Ye <redacted>
* remove visitor nonsense
* remove visitor comment
Signed-off-by: Xiaodong Ye <redacted>
* Address review comments
Signed-off-by: Xiaodong Ye <redacted>
---------
Signed-off-by: Xiaodong Ye <redacted>
Co-authored-by: slaren <redacted>
Georgi Gerganov [Fri, 4 Jul 2025 16:19:09 +0000 (19:19 +0300)]
metal : disable fast math in all quantize kernels (llama/14528)
ggml-ci
luyhcsu [Fri, 4 Jul 2025 03:50:07 +0000 (11:50 +0800)]
CANN: Replace aclrtMemsetSync with aclnnInplaceZero operator (llama/14002)
Co-authored-by: luyuhong <redacted>
Sigbjørn Skjæret [Thu, 3 Jul 2025 21:07:22 +0000 (23:07 +0200)]
ggml : implement GEGLU_ERF and GEGLU_QUICK ops (llama/14445)
lhez [Thu, 3 Jul 2025 18:22:24 +0000 (11:22 -0700)]
opencl : broadcast for soft_max (llama/14510)
Jeff Bolz [Thu, 3 Jul 2025 18:21:14 +0000 (13:21 -0500)]
vulkan: support mixed/deepseekR1 FA head sizes (llama/14509)
* vulkan: better parameterize FA by head sizes
* vulkan: support mixed/deepseekR1 FA head sizes
Johannes Gäßler [Thu, 3 Jul 2025 15:05:18 +0000 (17:05 +0200)]
ggml: backward pass for split swiglu (llama/14483)
Nicolò Scipione [Thu, 3 Jul 2025 09:00:03 +0000 (11:00 +0200)]
Fix conditional enabling following arch checks for ggml-sycl (llama/14504)
Signed-off-by: nscipione <redacted>
Georgi Gerganov [Thu, 3 Jul 2025 07:53:35 +0000 (10:53 +0300)]
kv-cache : use ggml_set_rows (llama/14285)
* kv-cache : use ggml_set_rows
ggml-ci
* graph : separate k and v indices
ggml-ci
* cont : remove redundant ifs
ggml-ci
* kv-cache : improve find_slot impl
* kv-cache : bounds-check when accessing slot_info indices
* kv-cache : add comments
ggml-ci
* ggml : add TODOs for adding GGML_OP_SET_ROWS support in the backends
ggml-ci
Georgi Gerganov [Thu, 3 Jul 2025 07:46:57 +0000 (10:46 +0300)]
ggml : fix FA mask dim 2 and 3 (llama/14505)
* ggml : fix FA mask dim 2 and 3
ggml-ci
* backends : unsupport batched FA in CUDA and Vulkan
ggml-ci
* vulkan : disable FA for mask->ne[2] != 1
Georgi Gerganov [Thu, 3 Jul 2025 04:48:32 +0000 (07:48 +0300)]
ggml : remove kompute backend (llama/14501)
ggml-ci
Aman Gupta [Wed, 2 Jul 2025 23:45:11 +0000 (07:45 +0800)]
CUDA: add dynamic shared mem to softmax, refactor general usage (llama/14497)
compilade [Wed, 2 Jul 2025 17:10:24 +0000 (13:10 -0400)]
llama : initial Mamba-2 support (llama/9126)
* llama : initial Mamba-2 support
* ggml : SIMD ggml_ssm_scan for Mamba-2
* ggml : improve ggml_mul speed when masking recurrent states
* llama : support running Mamba-Codestral-7B-v0.1
* llama : fix Mamba-2 conv state saving
* ggml : make the ggml_mul fast broadcast path more consistently formatted
* llama : remove unused variable
* llama : add missing break
* convert_hf : prefer SentencePiece tokenizer for Mamba-2 when present
The tokenzier.json of Mamba-Codestral-7B-v0.1 otherwise requires
workarounds to work correctly.
* llama : avoid redundant state copy for Mamba 1 and 2
* metal : attempt to adapt SSM_SCAN for Mamba-2
* metal : fix SSM_SCAN pipeline scope
* metal : use log and exp instead of log1pf and expf in SSM_SCAN
* metal : remove unused arguments for SSM_SCAN
The max index is 31, so trimming the arguments is necessary.
* metal : add back n_seqs to SSM_SCAN args
Whoops, this is needed for the offset in the concatenated output.
* metal : fix SSM_SCAN state head offset
* metal : fix wrong number of tokens per sequence in SSM_SCAN
* ggml : remove unused fast broadcast path in GGML_MUL
This was initially added because states were masked with ggml_mul,
but this is no longer done and so this "optimisation" is no longer
necessary, or at least not worth the additional code complexity.
* ggml : avoid multiply by D in GGML_OP_SSM_SCAN
This makes the weight buft detection in src/llama.cpp simpler.
* convert : transpose Mamba-2 A, D and reshape SSM_NORM
This breaks existing conversions of Mamba-2 models
to avoid some reshapes.
Not sure if it's a good idea,
but it makes the graph slightly cleaner.
* llama : more appropriate SSM_SCAN and SSM_CONV buft support checks
* convert : fix flake8 lint
* metal : fix confusion between ; and ,
* metal : add missing args for nb references in ssm_scan_f32_group
* metal : single-user mamba2 inference works
* kv-cache : remove const_cast when setting inputs for s_copy
And also fix multi-user inference for recurrent models
by using cell_id instead of i as the kv cell index
when populating s_copy.
* convert : avoid AutoConfig for Mamba and Mamba2 hparams
* kv-cache : allow context shift for recurrent models
* graph : fix recurrent state copies when avoiding copies
Works, but using lambda functions might not be that clean.
* ggml : fix mamba2 ssm scan when compiled with SVE
* ggml-cpu : reorder SVE FMA for consistency with other SIMD arches
* cuda : implement ssm scan for Mamba2
There is still room for improvement, but it works!
* cuda : adapt Mamba1 ssm scan to shape changes from Mamba2
* mamba : fix mismatched new and delete size for llm_build_mamba
Subclasses of llm_graph_context cannot have extra fields,
because the called destructor is not the one from the subclass.
This otherwise would cause problems when runnning Mamba-(1|2) inference
when compiled -DGGML_SANITIZE_ADDRESS=ON
* cuda : graceful fallback for Mamba-1 models with weird embd size
Aman Gupta [Wed, 2 Jul 2025 12:34:24 +0000 (20:34 +0800)]
CUDA: add softmax broadcast (llama/14475)
* CUDA: add softmax broadcast
* Pass by const ref
* Review: Use blockDims for indexing, remove designated initializers
* Add TODO for noncontigous input/output
Johannes Gäßler [Wed, 2 Jul 2025 11:42:12 +0000 (13:42 +0200)]
CUDA: broadcasting for FlashAttention mask (llama/14500)
Jeff Bolz [Tue, 1 Jul 2025 08:32:56 +0000 (03:32 -0500)]
vulkan: support softmax/FA batch and broadcast (llama/14449)
Georgi Gerganov [Sat, 12 Jul 2025 11:35:19 +0000 (14:35 +0300)]
sync : llama.cpp
Georgi Gerganov [Sat, 12 Jul 2025 11:33:49 +0000 (14:33 +0300)]
ggml : support bcast ggml_soft_max_ext, ggml_flash_attn_ext (llama/14435)
zhouwg [Wed, 2 Jul 2025 12:38:10 +0000 (20:38 +0800)]
opencl : fix possible buffer overflow in dump_tensor (llama/14490)
Eric Zhang [Wed, 2 Jul 2025 11:00:04 +0000 (19:00 +0800)]
opencl : skip empty nodes on cgraph compute (llama/14491)
lhez [Wed, 2 Jul 2025 07:07:42 +0000 (00:07 -0700)]
opencl : update upscale to support align corners (llama/14488)
Björn Ganster [Wed, 2 Jul 2025 05:19:31 +0000 (07:19 +0200)]
ggml : Callback before abort (llama/14481)
* Add a callback that will be called just before abort. This allows apps without a console to display a message to the user and save data if needed.
* Return previous callback to allow callback chaining
* style fixes
---------
Co-authored-by: Diego Devesa <redacted>
Georgi Gerganov [Tue, 1 Jul 2025 15:04:08 +0000 (18:04 +0300)]
ci : disable fast-math for Metal GHA CI (llama/14478)
* ci : disable fast-math for Metal GHA CI
ggml-ci
* cont : remove -g flag
ggml-ci
Chenguang Li [Tue, 1 Jul 2025 08:47:30 +0000 (16:47 +0800)]
CANN: update aclnnGroupedMatmulV2 to aclnnGroupedMatmulV3 (llama/14411)
* [CANN]update to aclnnGroupedMatmulV2
Signed-off-by: noemotiovon <redacted>
* Support MUL_MAT_ID on 310p
Signed-off-by: noemotiovon <redacted>
* fix editorconfig
Signed-off-by: noemotiovon <redacted>
---------
Signed-off-by: noemotiovon <redacted>
Jeff Bolz [Tue, 1 Jul 2025 08:43:08 +0000 (03:43 -0500)]
vulkan: Split large mul_mat_id to fit in shared memory (llama/14451)
Sigbjørn Skjæret [Tue, 1 Jul 2025 08:14:21 +0000 (10:14 +0200)]
add GELU_ERF (llama/14455)
Kai Pastor [Fri, 11 Jul 2025 14:47:57 +0000 (16:47 +0200)]
ci : simplify, switch to ninja (#1295)
* CI: Move GGML_N_THREADS to env
* CI: Move macos-13 into matrix
* CI: Build with ninja
* CI: Remove env
Kai Pastor [Thu, 10 Jul 2025 06:57:51 +0000 (08:57 +0200)]
examples : Test installed CMake config package (#1294)
* Add test-cmake example
* CI: Run test for installed cmake config
Acly [Thu, 3 Jul 2025 17:58:12 +0000 (19:58 +0200)]
vulkan : implement bilinear interpolation for ggml_upscale/ggml_interpolate (#1291)
* supports GGML_SCALE_MODE_BILINEAR and GGML_SCALE_FLAG_ALIGN_CORNERS
Acly [Thu, 3 Jul 2025 17:47:15 +0000 (19:47 +0200)]
vulkan : implement ggml_roll (#1290)
* vulkan : implement ggml_roll
* vulkan : refactor vk_op_unary_push_constants initialization
Daniel Bevenius [Wed, 2 Jul 2025 11:55:32 +0000 (13:55 +0200)]
ggml : add version function to get lib version (#1286)
* ggml : add version function to get lib version
This commit adds a function `ggml_version()` to the ggml library that
returns the version of the library as a string.
The motivation for this is that it can be useful to be able to
programmatically check the version of the ggml library being used.
Usage:
```c
printf("GGML version: %s\n", ggml_version());
```
Output:
```console
GGML version: 0.0.2219
```
* ggml : add ggml_commit()
---------
Co-authored-by: Georgi Gerganov <redacted>
Georgi Gerganov [Wed, 2 Jul 2025 05:07:23 +0000 (08:07 +0300)]
sync : whisper.cpp
Georgi Gerganov [Tue, 1 Jul 2025 08:10:25 +0000 (11:10 +0300)]
sync : llama.cpp
ggml-ci
Georgi Gerganov [Tue, 1 Jul 2025 08:05:48 +0000 (11:05 +0300)]
ggml : remove trailing whitespace (llama/0)
lhez [Tue, 1 Jul 2025 07:19:16 +0000 (00:19 -0700)]
opencl : add GEGLU, REGLU, SWIGLU (llama/14456)
Aman Gupta [Mon, 30 Jun 2025 15:57:04 +0000 (23:57 +0800)]
Add Conv2d for CPU (llama/14388)
* Conv2D: Add CPU version
* Half decent
* Tiled approach for F32
* remove file
* Fix tests
* Support F16 operations
* add assert about size
* Review: further formatting fixes, add assert and use CPU version of fp32->fp16
Georgi Gerganov [Mon, 30 Jun 2025 14:04:05 +0000 (17:04 +0300)]
metal : disable fast-math for some cpy kernels (llama/14460)
* metal : disable fast-math for some cpy kernels
ggml-ci
* cont : disable for q4_1
ggml-ci
* cont : disable for iq4_nl
ggml-ci
Romain Biessy [Mon, 30 Jun 2025 12:52:02 +0000 (14:52 +0200)]
ggml-cpu: sycl: Re-enable exp f16 (llama/14462)
Diego Devesa [Mon, 30 Jun 2025 10:43:15 +0000 (03:43 -0700)]
test-backend-ops : disable llama test (llama/14461)
xiaobing318 [Mon, 30 Jun 2025 09:48:24 +0000 (17:48 +0800)]
cmake : Remove redundant include path in CMakeLists.txt (llama/14452)
* Update docker.yml
修改docker.yml文件中的内容使其停止周期性的运行该workflow,如果想要运行该workflow可以手动启动
* Remove redundant include path in CMakeLists.txt
The parent directory '..' was removed from the include directories for the ggml-cpu-feats target, to avoid unnecessary include paths.
* Enable scheduled Docker image builds
Uncomments the workflow schedule to trigger daily Docker image rebuilds at 04:12 UTC, improving automation and keeping images up to date.
Vedran Miletić [Mon, 30 Jun 2025 08:17:18 +0000 (10:17 +0200)]
scripts : make the shell scripts cross-platform (llama/14341)
Akarshan Biswas [Sun, 29 Jun 2025 15:37:58 +0000 (21:07 +0530)]
SYCL: disable faulty fp16 exp kernel (llama/14395)
* SYCL: disable faulty fp16 CPU exponent for now
* Revert "SYCL: disable faulty fp16 CPU exponent for now"
This reverts commit
ed0aab1ec31b4eb4b0f275dd7acd41d96a375202 .
* SYCL: disable faulty fp16 CPU exponent for now
* Fix logic of disabling exponent kernel
Sigbjørn Skjæret [Sun, 29 Jun 2025 12:38:10 +0000 (14:38 +0200)]
ggml : fix unmerged GGML_FPxx_TO_FPxx refactoring (llama/14443)
Sigbjørn Skjæret [Sun, 29 Jun 2025 09:04:10 +0000 (11:04 +0200)]
ggml : implement REGLU/GEGLU/SWIGLU ops (llama/14158)
* implement unary REGLU/GEGLU/SWIGLU cpu ops
* relax constraints
* duplicate shape of source
* fix ggml_vec_geglu_f16
* special case gated ops
* implement unary REGLU/GEGLU/SWIGLU cuda ops
* tighten constraints again
* refactor into GGML_GLU_OP
* metal : add glu kernels
ggml-ci
* add CUDA_GLU_BLOCK_SIZE [no ci]
* more constraints and use 64bit ints
ggml-ci
* 64bit multiplication [no ci]
* implement swapped variants (cpu/cuda)
* update comment [no ci]
ggml-ci
* Vulkan: Add GLU ops and shaders
* SYCL: Implement fused kernel GEGLU, SWIGLU and REGLU for single up+gate
* ggml : implement GLU for split up/gate (llama/14181)
* implement GLU for split up/gate
* add tests for ggml_glu_split
* Vulkan: Implement glu_split logic and shader support
* add split to logging [no ci]
* SYCL: refactor element_size ops and add split up and gate support to gated kernels
* SYCL: switch GEGLU to use tanh approximation
---------
Co-authored-by: 0cc4m <redacted>
Co-authored-by: Akarshan <redacted>
* GGML: increase OP count in assertion
* Refactor: Optimize SYCL element-wise operations with unary function inlining
This commit refactors the SYCL element-wise operations to improve performance by:
- Inlining unary operations (sgn, abs, elu, gelu, silu, etc.) to reduce kernel launch overhead.
- Introducing helper functions `op_xxx` for each unary operation to encapsulate the logic.
- Replacing direct kernel calls with calls to these inlined functions.
- Using `__dpct_inline__` to encourage compiler inlining.
- Minor code cleanup and consistency improvements.
The changes aim to reduce kernel launch overhead and improve the overall efficiency of element-wise operations on SYCL devices.
* vulkan: Increase workgroup size for GLU, for performance (llama/14345)
* vulkan: Increase workgroup size for GLU, for performance
* vulkan: change GLU shaders to do one element per invocation rather than one row per workgroup
* merge fix
* metal : add support for split and swap
ggml-ci
---------
Co-authored-by: Georgi Gerganov <redacted>
Co-authored-by: 0cc4m <redacted>
Co-authored-by: Akarshan <redacted>
Co-authored-by: Jeff Bolz <redacted>
Jeff Bolz [Sun, 29 Jun 2025 07:43:36 +0000 (02:43 -0500)]
vulkan: Add fusion support for RMS_NORM+MUL (llama/14366)
* vulkan: Add fusion support for RMS_NORM+MUL
- Add a use_count to ggml_tensor, so we can detect if an output is used more than once.
- Change the ggml-vulkan rms_norm shader to optionally multiply by another tensor.
- Add detection logic and basic fusion logic in ggml-vulkan.
- Add some testing support for fusion. Rather than computing one node at a time, allow
for computing the whole graph and just testing one node's results. Add rms_norm_mul tests
and enable a llama test.
* extract some common fusion logic
* fix -Winconsistent-missing-override
* move ggml_can_fuse to a common function
* build fix
* C and C++ versions of can_fuse
* move use count to the graph to avoid data races and double increments when used in multiple threads
* use hash table lookup to find node index
* change use_counts to be indexed by hash table slot
* minimize hash lookups
style fixes
* last node doesn't need single use.
fix type.
handle mul operands being swapped.
* remove redundant parameter
---------
Co-authored-by: slaren <redacted>
Aman Gupta [Sat, 28 Jun 2025 17:30:53 +0000 (01:30 +0800)]
CUDA: add bf16 and f32 support to cublas_mul_mat_batched (llama/14361)
* CUDA: add bf16 and f32 support to cublas_mul_mat_batched
* Review: add type traits and make function more generic
* Review: make check more explicit, add back comments, and fix formatting
* Review: fix formatting, remove useless type conversion, fix naming for bools
Jeff Bolz [Sat, 28 Jun 2025 15:36:40 +0000 (10:36 -0500)]
vulkan: handle noncontig in the final case of ggml_vk_get_cpy_pipeline (llama/14378)
Jeff Bolz [Sat, 28 Jun 2025 15:17:09 +0000 (10:17 -0500)]
vulkan: lock accesses of pinned_memory vector (llama/14333)
Xinpeng Dou [Sat, 28 Jun 2025 09:35:41 +0000 (17:35 +0800)]
fix async_mode bug (llama/14432)
Jeff Bolz [Sat, 28 Jun 2025 03:35:30 +0000 (22:35 -0500)]
vulkan: Fix GGML_VULKAN_SHADER_DEBUG_INFO (llama/14427)
This setting needs to be passed through to vulkan-shaders-gen
Radoslav Gerganov [Fri, 27 Jun 2025 13:41:40 +0000 (16:41 +0300)]
ggml : add ggml_set_rows (llama/14274)
* ggml : add ggml_set_rows
Add ggml_set_rows(a, b, c) which copies rows from 'b' into 'a' using
indices from 'c'.
ref: #8366
* use I64 for indices
* ggml : add repeat impl for i64
* ggml : add ggml_is_contiguous_rows
* ggml : ggml_set_rows support broadcast
* ggml : ggml_set_rows support quantized dst
ggml-ci
* ggml : support GGML_TYPE_F32 ".from_float" trait
* ggml : ggml_set_rows update comment + better index name
* tests : add ggml_set_rows
* metal : add ggml_set_rows implementation
ggml-ci
* ggml : simplify forward_dup_f32
* ggml : fix supports_op
* tests : add comment to set_rows
* ggml : leave the repeat_i64 for a separate PR
ggml-ci
* ggml : set_rows use std::min instead of MIN
* ggml : better error message for set_rows unsupported type
* metal : perform op->type check only once
* tests : more consistent implementation + more tests
ggml-ci
---------
Co-authored-by: Georgi Gerganov <redacted>
bandoti [Thu, 26 Jun 2025 16:46:53 +0000 (13:46 -0300)]
cmake: regen vulkan shaders when shaders-gen sources change (llama/14398)
* Add shaders-gen sources as target deps
Georgi Gerganov [Thu, 26 Jun 2025 12:51:19 +0000 (15:51 +0300)]
metal : add special-case mat-vec mul for ne00 == 4 (llama/14385)
ggml-ci
Georgi Gerganov [Thu, 26 Jun 2025 12:50:15 +0000 (15:50 +0300)]
metal : batch rows copy in a single threadgroup (llama/14384)
* metal : batch rows copy in a single threadgroup
ggml-ci
* metal : handle some edge cases when threadgroup size is not a power of 2
ggml-ci
R0CKSTAR [Thu, 26 Jun 2025 04:11:59 +0000 (12:11 +0800)]
musa: enable fp16 mma (all) and cublas on qy2 (llama/13842)
* musa: enable fp16 mma (all) and cublas on qy2
Signed-off-by: Xiaodong Ye <redacted>
* Update src/ggml-cuda/ggml-cuda.cu
Co-authored-by: Johannes Gäßler <redacted>
* Address review comments
Signed-off-by: Xiaodong Ye <redacted>
* Address review comments
Signed-off-by: Xiaodong Ye <redacted>
* musa: disable MUL_MAT_ID (q2_k × f32) due to precision issues
Signed-off-by: Xiaodong Ye <redacted>
---------
Signed-off-by: Xiaodong Ye <redacted>
Co-authored-by: Johannes Gäßler <redacted>
Aaron Teo [Wed, 25 Jun 2025 21:49:04 +0000 (05:49 +0800)]
ggml-cpu: enable IBM NNPA Vector Intrinsics (llama/14317)
* ggml-cpu: add nnpa compile flag
Signed-off-by: Aaron Teo <redacted>
(cherry picked from commit
4a9f60c201573128f73a65999b3e5cc497fae5c1 )
* ggml-cpu: add fp16->fp32 nnpa first
Signed-off-by: Aaron Teo <redacted>
(cherry picked from commit
8d4a7987f9c1887f716be96250f2caeee0253929 )
* ggml-cpu: add fp32->fp16
Signed-off-by: Aaron Teo <redacted>
(cherry picked from commit
0ff0d6516247a41d2ade42b42cf0d676a4dd1627 )
* ggml-cpu: better variable names
Signed-off-by: Aaron Teo <redacted>
(cherry picked from commit
2f58bbcbb89c183340e252362b2a40651f573f1f )
* docs: update s390x docs
Signed-off-by: Aaron Teo <redacted>
(cherry picked from commit
01b929491b50071a5d0572235dcf5a449da70aa7 )
* ggml-cpu: add debugging prints to see if dlf16 is correct
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: fix print vs printf
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: fix float placeholder
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: ensure fp16 and fp32 load and stores are called
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: fp16 load ensured to hit
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: remove sigint from fp16 store
for some reason, the function is not getting a hit when debugged with
gdb. we will need to investigate further
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: activate nnpa for ggml_cpu_fp16_to_fp32
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: nnpa activate ggml_cpu_fp16_to_fp32 for 8 elements
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: nnpa switch to vec_xst test
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: switch to vec_xst for 4 element loops also
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: rework noop
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: remove noop, general code cleanup
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: clarify variable naming
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: activate nnpa for ggml_cpu_fp32_to_fp16
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: add breakpoint for debugging
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: test fix for conversion failure
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: disable fp32->fp16 nnpa conversions for now
there are some conversion failures in nnpa that requires the eyes of an
ibm stsm. will create a separate pr to introduce the fp32->fp16 change.
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: switch to elif macro
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: reattempt fp32->fp16
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: fix typo
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: reattempt fp32->fp16
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: fix compiler types
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: change to typedef vector types
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: add 4 element loops for fp32->fp16
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: clarified vector naming
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: bring back fp32->fp16 store nnpa
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: activate nnpa fp32->fp16 or fp16->fp32 compute
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: add nnpa macro check in ggml-impl
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: add missing __func__
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: diagnose why __NNPA__ macro is not being defined
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: import vecintrin.h to fix compiler errors
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: update macro tests
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: move s390x typedef to own header file
Signed-off-by: Aaron Teo <redacted>
* Revert "ggml-cpu: move s390x typedef to own header file"
This reverts commit
157f856c34589566151630e294563a420702db39 .
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: switch to importing ggml-cpu-impl instead
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: fix macro declaration
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: test more macros
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: add debug prints
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: bruteforce macro definitions
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: move macro definitions
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: add ggml-impl.h to cmakelists
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: switch to private macros
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: move s390x typedef to own header file
Signed-off-by: Aaron Teo <redacted>
(cherry picked from commit
157f856c34589566151630e294563a420702db39 )
* ggml-cpu: move things around
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: bring back compile macros
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: switch to quotes for import
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: add compiler error macro
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: add s390x detection in ggml-src
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: bring back compile definitions
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: undo cmakelists work
Signed-off-by: Aaron Teo <redacted>
* Revert "ggml-cpu: move s390x typedef to own header file"
This reverts commit
18d79e1a30b39d9aaa0bd58400c5cf2c32135c9a .
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: remove typedefs.h
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: remove typedef from cmakelists
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: add ggml-impl.h future notes
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: add todo comment for future reference
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: clarify naming of dlf16
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: remove unnecessary target compile definitions
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: move nnpa fp16->fp32 and fp32->fp16 to simd-mappings
Signed-off-by: Aaron Teo <redacted>
* ggml: refactor fp32->fp16 and fp16->fp32 simd to ggml-cpu
Signed-off-by: Aaron Teo <redacted>
* docs: update broken huggingface link for s390x
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: fix duplicate func names during compile
Signed-off-by: Aaron Teo <redacted>
* Revert "ggml-cpu: fix duplicate func names during compile"
This reverts commit
fbb733451f27677063b914d4f6c9a9841d45b38d .
Signed-off-by: Aaron Teo <redacted>
* Revert "ggml: refactor fp32->fp16 and fp16->fp32 simd to ggml-cpu"
This reverts commit
bd288e8fa52b5244f65cee21cb61062f1a9e0ca5 .
Signed-off-by: Aaron Teo <redacted>
* ggml: refactor fp16<->fp32 simd to ggml-cpu
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: fix missing simd-mappings.h import in quants.c
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: fix missing simd-mappings.h within repack
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: fix amx mmq missing simd-mappings.h
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: attempt at fixing loongarch failing build
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: move nnpa together with other fp16<->fp32 simd
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: fix wrong refactor of ggml-base
ref: https://github.com/ggml-org/llama.cpp/pull/14317#discussion_r2164176555
Signed-off-by: Aaron Teo <redacted>
* ggml: remove dependency on ggml-cpu from ggml-base
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: rename all fp16<->fp32 macros to prefix with ggml_cpu
ref: https://github.com/ggml-org/llama.cpp/pull/14317#discussion_r2164449406
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: remove mistaken fallback macro
fallback logic was already implemented but i was too sleepy to realise
Signed-off-by: Aaron Teo <redacted>
* ggml: move ggml_table_f32_f16 to ggml-cpu
ref: https://github.com/ggml-org/llama.cpp/pull/14317#discussion_r2164775006
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: move ggml_table_f32_f16 back to ggml-base due to ci failures
Signed-off-by: Aaron Teo <redacted>
* Revert "ggml-cpu: move ggml_table_f32_f16 back to ggml-base due to ci failures"
This reverts commit
32a3533564bdb7902cefb9c89b1c9e956a81ce29 .
Signed-off-by: Aaron Teo <redacted>
* Revert "ggml: move ggml_table_f32_f16 to ggml-cpu"
This reverts commit
9e40d984ad27d7b60392fb2b7548885201864fe4 .
Signed-off-by: Aaron Teo <redacted>
* ggml: move ggml_table_f32_f16 to ggml-cpu
ref: https://github.com/ggml-org/llama.cpp/pull/14317#discussion_r2164775006
Signed-off-by: Aaron Teo <redacted>
(cherry picked from commit
9e40d984ad27d7b60392fb2b7548885201864fe4 )
* ggml: move ggml_table_f32_f16 to ggml-cpu.c
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: extern c ggml_table_f32_f16 + chore docs
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: dedup ggml_table_f32_f16 from simd-mappings.h
we rely on the variable declaration in ggml-cpu.c instead
Signed-off-by: Aaron Teo <redacted>
* Revert "ggml-cpu: dedup ggml_table_f32_f16 from simd-mappings.h"
This reverts commit
f71b21d2f74f5e03ec0c2b4fefd3cbf395aecf16 .
Signed-off-by: Aaron Teo <redacted>
* ggml-cpu: bring back ggml_table_f32_f16
Signed-off-by: Aaron Teo <redacted>
* Revert "ggml-cpu: bring back ggml_table_f32_f16"
This reverts commit
2dce119178bed5ef5c8398c4230ddd14fef80e49 .
Signed-off-by: Aaron Teo <redacted>
* fix ggml time initialization
* fix f32_f16 table init
* remove extra line
---------
Signed-off-by: Aaron Teo <redacted>
Co-authored-by: slaren <redacted>
Sigbjørn Skjæret [Wed, 25 Jun 2025 21:26:51 +0000 (23:26 +0200)]
ggml : do not output unprintable characters on GGUF load failure (llama/14381)
Anton Mitkov [Wed, 25 Jun 2025 16:09:55 +0000 (17:09 +0100)]
sycl: GGML_SYCL_DISABLE_OPT on by default for all Intel Devices (llama/13973)
lhez [Tue, 24 Jun 2025 18:46:25 +0000 (11:46 -0700)]
opencl: ref count `ggml_backend_opencl_context` and refactor profiling (llama/14254)
* Move profiling info into `ggml_backend_opencl_context`
* Add `enqueue_ndrange_kernel` to launch kernel
uvos [Mon, 23 Jun 2025 23:12:56 +0000 (01:12 +0200)]
CUDA/HIP: optimize mmv paths taken for HIP devices (llama/14324)
Co-authored-by: Johannes Gäßler <redacted>
Johannes Gäßler [Mon, 23 Jun 2025 11:11:31 +0000 (13:11 +0200)]
CUDA: mul_mat_v support for batch sizes > 1 (llama/14262)
* CUDA: mul_mat_v support for batch sizes > 1
* use 64 bit math for initial offset calculation
uvos [Sun, 22 Jun 2025 14:51:23 +0000 (16:51 +0200)]
HIP: enable vec fattn on RDNA4 (llama/14323)
Aman Gupta [Sun, 22 Jun 2025 04:39:54 +0000 (12:39 +0800)]
CUDA: add mean operation (llama/14313)
* CUDA: add mean operation
* add back sum_rows_f32_cuda
* Review: early exit if col!=0
Markus Tavenrath [Sat, 21 Jun 2025 06:17:12 +0000 (08:17 +0200)]
Add support for VK_EXT_debug_utils to add labels to Vulkan objects. (llama/13792)
* Add support for VK_EXT_debug_utils to add labels to Vulkan objects. In step 1 compute pipelines are getting labeled.
* remove #ifdef for debug utils and add queue marker.
Georgi Gerganov [Sat, 21 Jun 2025 05:04:18 +0000 (08:04 +0300)]
metal : fix thread-safety (llama/14300)
ggml-ci
Acly [Tue, 1 Jul 2025 07:11:00 +0000 (09:11 +0200)]
ggml-cpu : "align corners" for bilinear upscale/downscale (#1285)
* add "align corners" mode for bilinear upscale, and allow downscaling
* add ggml_interpolate, deprecate ggml_upscale_ext, pass in align-corners as bit-flag
* test-backend-ops: replace ggml_upscale_ext with ggml_interpolate, add test cases for downscale and align-corners
Acly [Wed, 25 Jun 2025 10:16:22 +0000 (12:16 +0200)]
build : fix build with clang-cl on Windows (#1284)
* build : fix building tests with clang-cl on Windows
- clang-cl.exe (clang with MSVC CLI) doesn't like the space in /STACK option
- cl.exe (MSVC) works either way
* build : fix MSVC compiler warnings in test-roll.cpp
Daniel Bevenius [Tue, 24 Jun 2025 04:10:16 +0000 (06:10 +0200)]
ggml-quants : rename best_mad to best_error (#1283)
This commit renames the variable `best_mad` to `best_error` in the
`make_qkx2_quants` function.
The motivation for this is that the name `best_mad` can be somewhat
confusing if mean absolute deviation (MAD) is not in use.
Georgi Gerganov [Sat, 21 Jun 2025 06:21:28 +0000 (09:21 +0300)]
tests : cleanup old tests (#1282)
ggml-ci
Georgi Gerganov [Fri, 20 Jun 2025 18:04:04 +0000 (21:04 +0300)]
sync : llama.cpp
ggml-ci
Aman Gupta [Fri, 20 Jun 2025 14:48:24 +0000 (22:48 +0800)]
CUDA: add conv_2d_transpose (llama/14287)
* CUDA: add conv_2d_transpose
* remove direct include of cuda_fp16
* Review: add brackets for readability, remove ggml_set_param and add asserts
Nicolò Scipione [Fri, 20 Jun 2025 13:07:21 +0000 (15:07 +0200)]
sycl: add usage of enqueue_functions extension (llama/14244)
* Add header and namespace to use enqueue_functions extension
* Convert submit and parallel_for to use new extension in convert.cpp
* Convert submit and parallel_for to use extension in ggml-sycl.cpp
* Convert submit and parallel_for to use extension in gla.cpp
* Convert submit and parallel_for in mmq.cpp
* Convert submit and parallel_for in mmvq.cpp
* Convert submit and parallel_for in remaining files
* Convert all simple parallel_for to nd_launch from enqueue_functions
extension
* Wrapping extension in general function
Create a general function that enable the enqueue_functions extension if
it is enable in the compiler, otherwise call the general SYCL function
to launch kernels.
---------
Signed-off-by: nscipione <redacted>
Christian Kastner [Fri, 20 Jun 2025 12:17:32 +0000 (12:17 +0000)]
Implement GGML_CPU_ALL_VARIANTS for PowerPC (llama/14286)
* Add PowerPC feature detection and scoring
* ggml-cpu: Implement GGML_CPU_ALL_VARIANTS for PowerPC
* ggml-cpu: Delay some initializations until function is called
When using GGML_BACKEND_DL=ON, these initializations might use
instructions that are not supported by the current CPU.
---------
Co-authored-by: Diego Devesa <redacted>
Diego Devesa [Fri, 20 Jun 2025 11:57:36 +0000 (04:57 -0700)]
cuda : synchronize graph capture and cublas handle destruction (llama/14288)
Workarounds an issue that may cause CUDA graph capture to fail when a cuBLAS handle is destroyed in a different thread
Georgi Gerganov [Fri, 20 Jun 2025 08:19:15 +0000 (11:19 +0300)]
ggml : fix repack work size for mul_mat_id (llama/14292)
ggml-ci
Charles Xu [Fri, 20 Jun 2025 07:51:01 +0000 (09:51 +0200)]
ggml: Update KleidiAI to v1.9.0 (llama/14277)
Aman Gupta [Fri, 20 Jun 2025 01:50:24 +0000 (09:50 +0800)]
CUDA: add conv_2d_dw (llama/14265)
* CUDA: add conv_2d_dw
* better naming
* simplify using template
* Review: fix operation ordering in ggml-cuda, use __forceinline__, use more const