]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/log
pkg/ggml/sources/llama.cpp
3 months agoCUDA: Fix clang warnings (#12540)
R0CKSTAR [Mon, 24 Mar 2025 10:28:34 +0000 (18:28 +0800)]
CUDA: Fix clang warnings (#12540)

Signed-off-by: Xiaodong Ye <redacted>
3 months agommap : skip resource limit checks on AIX (#12541)
Prajwal B Mehendarkar [Mon, 24 Mar 2025 10:17:10 +0000 (15:47 +0530)]
mmap : skip resource limit checks on AIX (#12541)

3 months agovulkan: fix mul_mat_vec failure in backend tests (#12529)
Jeff Bolz [Mon, 24 Mar 2025 06:56:17 +0000 (01:56 -0500)]
vulkan: fix mul_mat_vec failure in backend tests (#12529)

The OOB calculation could be wrong if the last iteration was during one of
the unrolled loops. Adjust the unrolling counts to avoid this. Add a couple
new backend tests that hit this failure on NVIDIA GPUs.

3 months agoserver : Add verbose output to OAI compatible chat endpoint. (#12246)
Marius Gerdes [Sun, 23 Mar 2025 18:30:26 +0000 (19:30 +0100)]
server : Add verbose output to OAI compatible chat endpoint. (#12246)

Add verbose output to server_task_result_cmpl_final::to_json_oaicompat_chat_stream, making it conform with server_task_result_cmpl_final::to_json_oaicompat_chat, as well as the other to_json methods.

3 months agoinstall : add macports (#12518)
Lars Sonchocky-Helldorf [Sun, 23 Mar 2025 08:21:48 +0000 (09:21 +0100)]
install : add macports (#12518)

MacPorts section added

3 months agollama : gemma3 : use output tensor if it exists in model weight (#12506)
Xuan-Son Nguyen [Sat, 22 Mar 2025 22:28:19 +0000 (23:28 +0100)]
llama : gemma3 : use output tensor if it exists in model weight (#12506)

* llama : gemma3 : use output tensor if it exists in model weight

* also add to the llm_tensor_names

3 months agoggml : fix quantized cpy op (#12310)
Georgi Gerganov [Sat, 22 Mar 2025 14:23:26 +0000 (16:23 +0200)]
ggml : fix quantized cpy op (#12310)

* ggml : fix quantized cpy op

ggml-ci

* tests : add cpy tests for all types

ggml-ci

* tests : add BF16 copy tests

ggml-ci

* tests : fix loop for same-type copy

ggml-ci

* tests : add option to permute the dst tensor

ggml-ci

3 months agomusa: refine compute capability (#12493)
R0CKSTAR [Sat, 22 Mar 2025 09:11:37 +0000 (17:11 +0800)]
musa: refine compute capability (#12493)

* musa: refine compute capability

Signed-off-by: Xiaodong Ye <redacted>
* Address review comments

Signed-off-by: Xiaodong Ye <redacted>
---------

Signed-off-by: Xiaodong Ye <redacted>
3 months agovulkan: Optimize mul_mat_vec p021 and nc shaders (#12505)
Jeff Bolz [Sat, 22 Mar 2025 08:40:11 +0000 (03:40 -0500)]
vulkan: Optimize mul_mat_vec p021 and nc shaders (#12505)

* tests: add mul_mat perf/functional tests for p021/nc vulkan shaders

* vulkan: Optimize mul_mat_vec p021 and nc shaders.

These shaders are used in attention calculations, and when the KV cache grows
large they start to dominate the run time. For the nc shader (which is called
with large 'k' dimension), use unrolling and vector loads. For the p021 shader
(which is called with large 'm' and small 'k' dimensions), take advantage of
grouped query attention to reuse loads from the A matrix for the whole group,
and reduce the number of workgroups (too much overhead from tiny dispatches).

Using subgroupAdd in the p021 shader also helps, use that conditionally.

3 months agoVulkan: RTE rounding for cpy to quant (#12480)
stduhpf [Fri, 21 Mar 2025 19:34:50 +0000 (20:34 +0100)]
Vulkan: RTE rounding for cpy to quant (#12480)

* Vulkan: RTE rounding for cpy to quant

Co-Authored-By: Jeff Bolz <redacted>
* remove trailing whitespace

* avoid duplicating pipeline_cpy_f32_quant

* fix copypasting issue

* remove duplicated code

---------

Co-authored-by: Jeff Bolz <redacted>
3 months agovulkan: workaround for AMD Windows driver 16 bit unpack8 bug (#12472)
Eve [Fri, 21 Mar 2025 19:27:47 +0000 (19:27 +0000)]
vulkan: workaround for AMD Windows driver 16 bit unpack8 bug (#12472)

3 months agomodel : do not repack if a GPU device is present (#12498)
Georgi Gerganov [Fri, 21 Mar 2025 14:14:29 +0000 (16:14 +0200)]
model : do not repack if a GPU device is present (#12498)

ggml-ci

3 months agochore : cleanup llama_model_loader::TENSOR_ usage (#12492)
Sigbjørn Skjæret [Fri, 21 Mar 2025 09:21:36 +0000 (10:21 +0100)]
chore : cleanup llama_model_loader::TENSOR_ usage (#12492)

3 months agollama-tts : avoid crashes related to bad model file paths (#12482)
marcoStocchi [Fri, 21 Mar 2025 09:12:45 +0000 (10:12 +0100)]
llama-tts : avoid crashes related to bad model file paths (#12482)

3 months ago[SYCL] Fix build on Windows when ccache enabled (#9954) (#9976)
蕭澧邦 [Fri, 21 Mar 2025 06:58:47 +0000 (14:58 +0800)]
[SYCL] Fix build on Windows when ccache enabled (#9954) (#9976)

* [SYCL] Fix build on Windows when ccache enabled (#9954)

* take effect only on windows and force it to icl

---------

Co-authored-by: Romain Biessy <redacted>
3 months agosycl: cleanup oneDNN related code (#12097)
Svetlozar Georgiev [Fri, 21 Mar 2025 02:15:56 +0000 (02:15 +0000)]
sycl: cleanup oneDNN related code (#12097)

3 months agowebui : Prevent rerendering on textarea input (#12299)
Woof Dog [Thu, 20 Mar 2025 14:57:43 +0000 (14:57 +0000)]
webui : Prevent rerendering on textarea input (#12299)

* webui: Make textarea uncontrolled to eliminate devastating lag

* Update index.html.gz

* use signal-style implementation

* rm console log

* no duplicated savedInitValue set

---------

Co-authored-by: Xuan Son Nguyen <redacted>
3 months agollama : make Qwen2MoE QKV bias optional (#12477)
Sigbjørn Skjæret [Thu, 20 Mar 2025 11:49:59 +0000 (12:49 +0100)]
llama : make Qwen2MoE QKV bias optional (#12477)

3 months agoggml : block interleaving support for Q4_K quantization for x86 AVX2 architecture...
Srihari-mcw [Thu, 20 Mar 2025 11:35:34 +0000 (17:05 +0530)]
ggml : block interleaving support for Q4_K quantization for x86 AVX2 architecture (#12332)

* Add block interleaving support for Q4_K quantization

* Remove whitespaces and fix CI/CD issues

* Update pointer of bsums from int16_t to const int16_t

* Add vector version of quantize_q8_K_4x8 function

* Update code formatting based on review comments

3 months agoconvert : avoid calls to tokenizer.added_tokens_decoder (#12473)
Bartowski [Thu, 20 Mar 2025 06:36:37 +0000 (02:36 -0400)]
convert : avoid calls to tokenizer.added_tokens_decoder (#12473)

tokenizer.added_tokens_decoder returns a fresh dict every time relatively slowly (~0.04s on average) which results in massive slowdowns when we have a huge number of added tokens

3 months agocontext : clear sets containing encoder output sequence ids before storing new values...
fairydreaming [Wed, 19 Mar 2025 20:01:57 +0000 (21:01 +0100)]
context : clear sets containing encoder output sequence ids before storing new values (#12470)

Co-authored-by: Stanisław Szymczyk <redacted>
3 months agoCUDA: Improve flash decoding kernel GPU occupancy for BS=1 case (#12183)
Gaurav Garg [Wed, 19 Mar 2025 19:52:06 +0000 (01:22 +0530)]
CUDA: Improve flash decoding kernel GPU occupancy for BS=1 case (#12183)

- Find out active blocks per SM using cudaOccupancyMaxActiveBlocksPerMultiprocessor API. Use this value to determine the optimal parallel_blocks value.
- Prefer vector flash attention kernels over MMA kernel for BS=1

Fixes Issue: #12182
---------

Co-authored-by: Johannes Gäßler <redacted>
3 months agovulkan: optimize iq1 coopmat2 dequant functions (#12427)
Jeff Bolz [Wed, 19 Mar 2025 18:56:23 +0000 (13:56 -0500)]
vulkan: optimize iq1 coopmat2 dequant functions (#12427)

3 months agoFix visionOS build and add CI (#12415)
Guus Waals [Wed, 19 Mar 2025 10:15:23 +0000 (10:15 +0000)]
Fix visionOS build and add CI (#12415)

* ci: add visionOS build workflow

Add a new GitHub Actions workflow for building on visionOS with CMake and Xcode.

* ggml: Define _DARWIN_C_SOURCE for visionOS to fix missing u_xxx typedefs

* ci: remove define hacks for u_xxx system types

---------

Co-authored-by: Giovanni Petrantoni <redacted>
3 months agollama : add support for GPT2, Bloom and CodeShell tied word embeddings (#12456)
Sigbjørn Skjæret [Wed, 19 Mar 2025 08:08:49 +0000 (09:08 +0100)]
llama : add support for GPT2, Bloom and CodeShell tied word embeddings (#12456)

* Add support for GPT2, Bloom and CodeShell tied word embeddings

* Deduplicate tied word embeddings weights

* Workaround for incorrect weight map

It appears transformer.wte.weight is in the weight map even though the weights are not there, remove it if output weights are encountered first.

* check++

* fatfingers--

3 months agoconvert : Support chat_template.json (#12460)
Sigbjørn Skjæret [Wed, 19 Mar 2025 07:58:13 +0000 (08:58 +0100)]
convert : Support chat_template.json (#12460)

3 months agovulkan: Submit once enough matmul work has been recorded (#12406)
Jeff Bolz [Wed, 19 Mar 2025 07:26:26 +0000 (02:26 -0500)]
vulkan: Submit once enough matmul work has been recorded (#12406)

I've been seeing significantly worse performance for tg with flash attention
enabled vs disabled, and it seems to be related to the submit heuristic.
Change the heuristic to check how many bytes worth of weight matrix are
used and flush every 100MB, and ramp up after the first few submits.
This seems to resolve the issue, and also increases perf for non-FA a bit.

3 months agoopencl: improve profiling (#12442)
lhez [Tue, 18 Mar 2025 19:54:55 +0000 (12:54 -0700)]
opencl: improve profiling (#12442)

* opencl: more profiling timing

* opencl: generate trace for profiling

* opencl: reduce profiling overhead

* Populate profiling timing info at the end rather than after each
  kernel run

* opencl: fix for chrome tracing

3 months agograph : normalize Q, K, V shapes + sync cross attention (#12449)
Georgi Gerganov [Tue, 18 Mar 2025 19:35:19 +0000 (21:35 +0200)]
graph : normalize Q, K, V shapes + sync cross attention (#12449)

* graph : normalize Q, K, V shapes and add comments

ggml-ci

* context : synchronize before getting cross attention data

* model : fix command-r attention norm check

3 months agomusa: override warp_size of musa device to 32 (#12445)
R0CKSTAR [Tue, 18 Mar 2025 18:28:26 +0000 (02:28 +0800)]
musa: override warp_size of musa device to 32 (#12445)

Signed-off-by: Xiaodong Ye <redacted>
3 months agollama : support converting Mistral Small text-only (#12450)
Xuan-Son Nguyen [Tue, 18 Mar 2025 18:16:19 +0000 (19:16 +0100)]
llama : support converting Mistral Small text-only (#12450)

3 months agospeculative : fix seg fault in certain cases (#12454)
Georgi Gerganov [Tue, 18 Mar 2025 17:35:11 +0000 (19:35 +0200)]
speculative : fix seg fault in certain cases (#12454)

3 months agollama : add support for EXAONE tied word embeddings (#12451)
Xuan-Son Nguyen [Tue, 18 Mar 2025 16:24:33 +0000 (17:24 +0100)]
llama : add support for EXAONE tied word embeddings (#12451)

3 months agocontext : always use non-causal attention for encoder graphs (#12447)
Georgi Gerganov [Tue, 18 Mar 2025 11:05:49 +0000 (13:05 +0200)]
context : always use non-causal attention for encoder graphs (#12447)

* context : always use non-causal attention for encoder graphs

ggml-ci

* context : move the change to llama_context::encode()

ggml-ci

3 months agoSYCL: using graphs is configurable by environment variable and compile option (#12371)
Łukasz Ślusarczyk [Tue, 18 Mar 2025 10:16:31 +0000 (11:16 +0100)]
SYCL: using graphs is configurable by environment variable and compile option (#12371)

* alberto changes

* enable sycl graphs by env variable

* fixed compilation warnings in ggml-sycl.cpp

* renamed graph variables

* fix markdown in docs/backend/SYCL.md

Co-authored-by: Romain Biessy <redacted>
* fix markdown in docs/backend/SYCL.md again

* compiling graphs by default, renamed graph_enable to graph_disable

---------

Co-authored-by: Romain Biessy <redacted>
3 months agoserver : fix warmup draft cache type (#12446)
Georgi Gerganov [Tue, 18 Mar 2025 10:05:42 +0000 (12:05 +0200)]
server : fix warmup draft cache type (#12446)

ggml-ci

3 months agocmake : fix PowerPC build (#12241)
Prajwal B Mehendarkar [Tue, 18 Mar 2025 09:37:33 +0000 (15:07 +0530)]
cmake : fix PowerPC build (#12241)

Closes #12240

3 months agoggml : add SVE support for q6_K_q8_K (#12361)
fj-y-saito [Tue, 18 Mar 2025 08:14:39 +0000 (17:14 +0900)]
ggml : add SVE support for q6_K_q8_K (#12361)

3 months agoVulkan: Default to 1GB allocations instead of 4GB to avoid fragmentation and driver...
0cc4m [Tue, 18 Mar 2025 06:21:40 +0000 (07:21 +0100)]
Vulkan: Default to 1GB allocations instead of 4GB to avoid fragmentation and driver issues (#12434)

3 months agofixed compilation warnings in ggml-sycl (#12424)
Łukasz Ślusarczyk [Tue, 18 Mar 2025 00:51:25 +0000 (01:51 +0100)]
fixed compilation warnings in ggml-sycl (#12424)

3 months agollama: Add support for RWKV v7 architecture (#12412)
Molly Sophia [Mon, 17 Mar 2025 23:27:50 +0000 (07:27 +0800)]
llama: Add support for RWKV v7 architecture (#12412)

* ggml: Add op l2_norm

Signed-off-by: Molly Sophia <redacted>
* ggml: Add op rwkv_wkv7

Signed-off-by: Molly Sophia <redacted>
* llama: Add support for RWKV7 and ARWKV7 models

Signed-off-by: Molly Sophia <redacted>
* llama: fix inference with RWKV6Qwen2

Signed-off-by: Molly Sophia <redacted>
* llama: add more (a)rwkv7 variants in size

Signed-off-by: Molly Sophia <redacted>
* Apply code-format changes

Signed-off-by: Molly Sophia <redacted>
* fix MUSA build

Signed-off-by: Molly Sophia <redacted>
* llama: fix shape error with rwkv using llama-parallel

Signed-off-by: Molly Sophia <redacted>
---------

Signed-off-by: Molly Sophia <redacted>
3 months agodocs : bring llama-cli conversation/template docs up-to-date (#12426)
Sigbjørn Skjæret [Mon, 17 Mar 2025 20:14:32 +0000 (21:14 +0100)]
docs : bring llama-cli conversation/template docs up-to-date (#12426)

3 months agocuda : enable CUDA Graph on CUDA Toolkit < 12.x (#12394)
Gaurav Garg [Mon, 17 Mar 2025 18:25:13 +0000 (23:55 +0530)]
cuda : enable CUDA Graph on CUDA Toolkit < 12.x (#12394)

* Enable CUDA Graph on CTK < 12.x

`cudaGraphExecUpdate` API was changed on 12.x. For this reason CUDA graph support was disabled on older CUDA toolkit. This change enables CUDA support in CTK version < 12.x by using older API if CTK < 12.x.

* Fix compilation errors with MUSA

* Disable CUDA Graph for MUSA

3 months agoggml-vulkan: remove unused find_program(glslc) (#12416)
Guus Waals [Mon, 17 Mar 2025 16:35:43 +0000 (00:35 +0800)]
ggml-vulkan: remove unused find_program(glslc) (#12416)

It's already found by FindVulkan.cmake in the parent CMakeLists

3 months agovulkan: Add N/2 and N/4 optimized paths in coopmat2 shader (#12312)
Jeff Bolz [Mon, 17 Mar 2025 14:26:18 +0000 (09:26 -0500)]
vulkan: Add N/2 and N/4 optimized paths in coopmat2 shader (#12312)

3 months agovulkan: subgroup size tuning (#12087)
Daniele [Mon, 17 Mar 2025 11:42:33 +0000 (12:42 +0100)]
vulkan: subgroup size tuning (#12087)

* vulkan: subgroup size test

* Vulkan: Add device architecture enum and logic to recognize AMD generations

* vulkan: use new architecture logic to specify subgroup size

* Initial vulkan subgroup size tuning for RDNA3

* vulkan: commonize RDNA subgroup tuning

* vulkan: override subgroup size if required_subgroup_size = 0

* vulkan: disable warp 32 for RDNA3

* vulkan: fine tuned RDNA1 subgroup sizes

* vulkan: adjusted subgroup size map

* vulkan: fixed RDNA2 subgroup map

---------

Co-authored-by: 0cc4m <redacted>
3 months agovulkan: use fp32 in coopmat2 q4_k dequant function (#12309)
Jeff Bolz [Mon, 17 Mar 2025 09:43:35 +0000 (04:43 -0500)]
vulkan: use fp32 in coopmat2 q4_k dequant function (#12309)

3 months agovulkan: Pad N dimension of B matrix for coopmat2 perf, to avoid bounds checking ...
Jeff Bolz [Mon, 17 Mar 2025 09:41:59 +0000 (04:41 -0500)]
vulkan: Pad N dimension of B matrix for coopmat2 perf, to avoid bounds checking (#12273)

* vulkan: Pad N dimension of B matrix for coopmat2 perf, to avoid bounds checking

3 months agovulkan: Adjust coopmat2 tile sizes and selection heuristic (#12258)
Jeff Bolz [Mon, 17 Mar 2025 09:35:00 +0000 (04:35 -0500)]
vulkan: Adjust coopmat2 tile sizes and selection heuristic (#12258)

3 months agocmake : enable building llama.cpp using system libggml (#12321)
Christian Kastner [Mon, 17 Mar 2025 09:05:23 +0000 (10:05 +0100)]
cmake : enable building llama.cpp using system libggml (#12321)

* cmake: Factor out compiler flag function from ggml

llama.cpps's build requires it, too, and we may want to make use of it
without add_subdirectory(ggml).

* cmake: Enable building against system ggml

This facilitates package maintenance for Linux distributions, where the
libggml library most likely will be shipped as an individual package
upon which a llama.cpp package depends.

3 months agoSYCL: set extras only on GGML_TYPE_Q4_0 (#12366)
Akarshan Biswas [Mon, 17 Mar 2025 01:45:12 +0000 (07:15 +0530)]
SYCL: set extras only on GGML_TYPE_Q4_0 (#12366)

* SYCL: set extras only on GGML_TYPE_Q4_0

* release tensor_extras in reset buffer interface

3 months agollama : fix OLMo-2-0325-32B-Instruct K-norm size (#12400)
Sigbjørn Skjæret [Sun, 16 Mar 2025 17:46:36 +0000 (18:46 +0100)]
llama : fix OLMo-2-0325-32B-Instruct K-norm size (#12400)

3 months agocontext : fix init of n_outputs (#12397)
Georgi Gerganov [Sun, 16 Mar 2025 17:29:36 +0000 (19:29 +0200)]
context : fix init of n_outputs (#12397)

ggml-ci

3 months agoci : add --symlinks to xcframework zip command (#12409)
Daniel Bevenius [Sun, 16 Mar 2025 17:22:05 +0000 (18:22 +0100)]
ci : add --symlinks to xcframework zip command (#12409)

This commit adds the --symlinks option to the zip command used to create
the xcframework zip file. This is necessary to create symlinks in the
zip file. Without this option,  the Versions symlink is stored as a
regular directory entry in the zip file, rather than as a symlink in the
zip which causes the followig error in xcode:
```console
Couldn't resolve framework symlink for '/Users/danbev/work/ai/llama.cpp/tmp_1/build-apple/llama.xcframework/macos-arm64_x86_64/llama.framework/Versions/Current': readlink(/Users/danbev/work/ai/llama.cpp/tmp_1/build-apple/llama.xcframework/macos-arm64_x86_64/llama.framework/Versions/Current): Invalid argument (22)
```

Refs: https://github.com/ggml-org/llama.cpp/pull/11996#issuecomment-2727026377

3 months agollama-tts : add '-o' option (#12398)
marcoStocchi [Sat, 15 Mar 2025 16:23:11 +0000 (17:23 +0100)]
llama-tts : add '-o' option (#12398)

* added -o option to specify an output file name

* llama-tts returns ENOENT in case of file write error

note : PR #12042 is closed as superseded with this one.

3 months agoSYCL: Delete redundant plus sign and space (#12391)
aubreyli [Sat, 15 Mar 2025 14:49:03 +0000 (22:49 +0800)]
SYCL: Delete redundant plus sign and space (#12391)

3 months agoSYCL : support non-contiguous tensors in binary ops (add, sub, etc) (#12399)
fairydreaming [Sat, 15 Mar 2025 14:19:30 +0000 (15:19 +0100)]
SYCL : support non-contiguous tensors in binary ops (add, sub, etc) (#12399)

* sycl : support non-contiguous tensors in binary ops

* sycl : silence unused variable warning

---------

Co-authored-by: Stanisław Szymczyk <redacted>
3 months ago[CANN]MUL_MAT optimization (#12382)
Chenguang Li [Sat, 15 Mar 2025 01:31:08 +0000 (09:31 +0800)]
[CANN]MUL_MAT optimization (#12382)

3 months agoAdd CLI arg to llama-run to adjust the number of threads used (#12370)
Eric Curtin [Fri, 14 Mar 2025 16:41:20 +0000 (16:41 +0000)]
Add CLI arg to llama-run to adjust the number of threads used (#12370)

We default to 4, sometimes we want to manually adjust this

Signed-off-by: Eric Curtin <redacted>
3 months agomain : add -sysf / --system-prompt-file (#12249) (#12250)
Sigbjørn Skjæret [Fri, 14 Mar 2025 15:57:05 +0000 (16:57 +0100)]
main : add -sysf / --system-prompt-file (#12249) (#12250)

* add system_prompt_file

* add -sysf / --system-prompt-file

* remove system_prompt_file

3 months agoLoad all MoE experts during warmup (#11571)
fairydreaming [Fri, 14 Mar 2025 12:47:05 +0000 (13:47 +0100)]
Load all MoE experts during warmup (#11571)

* llama : introduce llama_set_warmup() API call that controls warmup mode; use all MoE experts during warmup

* common : use new API to enable warmup mode during model warmup

---------

Co-authored-by: Stanisław Szymczyk <redacted>
3 months agoserver: fix "--grammar-file" parameter (#12285)
Victor [Fri, 14 Mar 2025 10:21:17 +0000 (11:21 +0100)]
server: fix "--grammar-file" parameter (#12285)

3 months agograph : simplify attn input build for unified KV cache (#12381)
Georgi Gerganov [Fri, 14 Mar 2025 08:47:44 +0000 (10:47 +0200)]
graph : simplify attn input build for unified KV cache (#12381)

ggml-ci

3 months agohparams : add SWA rope parameters (#12374)
Georgi Gerganov [Fri, 14 Mar 2025 07:03:24 +0000 (09:03 +0200)]
hparams : add SWA rope parameters (#12374)

ggml-ci

3 months agollama : fix Gemma3 SWA KV cache shift (#12373)
Georgi Gerganov [Thu, 13 Mar 2025 17:08:07 +0000 (19:08 +0200)]
llama : fix Gemma3 SWA KV cache shift (#12373)

* llama : fix Gemma3 SWA KV cache shift

ggml-ci

* hparams : add comment [no ci]

3 months agoarg : no n_predict = -2 for examples except for main and infill (#12364)
Xuan-Son Nguyen [Thu, 13 Mar 2025 11:34:54 +0000 (12:34 +0100)]
arg : no n_predict = -2 for examples except for main and infill (#12364)

3 months agollama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)
Georgi Gerganov [Thu, 13 Mar 2025 10:35:44 +0000 (12:35 +0200)]
llama : refactor llama_context, llama_kv_cache, llm_build_context (#12181)

* llama : refactor llama_context, llama_kv_cache, llm_build_context

ggml-ci

* graph : don't mutate the KV cache during defrag

ggml-ci

* context : reduce virtuals + remove test function

ggml-ci

* context : move interface implementation to source file + factory

ggml-ci

* graph : move KV cache build functions to llama_context impl

ggml-ci

* graph : remove model reference from build_pooling

ggml-ci

* graph : remove llama_model reference

ggml-ci

* kv_cache : provide rope factors

ggml-ci

* graph : rework inputs to use only unique_ptr, remove attn input abstraction

ggml-ci

* context : remove llama_context_i abstraction

ggml-ci

* context : clean-up

ggml-ci

* graph : clean-up

ggml-ci

* llama : remove redundant keywords (struct, enum)

ggml-ci

* model : adapt gemma3

ggml-ci

* graph : restore same attention ops as on master

ggml-ci

* llama : remove TODO + fix indent

ggml-ci

3 months agoserver : fix crash when using verbose output with input tokens that are not in printa...
Ishaan Gandhi [Thu, 13 Mar 2025 10:10:05 +0000 (06:10 -0400)]
server : fix crash when using verbose output with input tokens that are not in printable range (#12178) (#12338)

* Fix DOS index bug

* Remove new APIs

* remove extra line

* Remove from API

* Add extra newline

* Update examples/server/server.cpp

---------

Co-authored-by: Xuan-Son Nguyen <redacted>
3 months agoUpdate build.yml for Windows Vulkan builder to use Vulkan 1.4.304 SDK for VK_NV_coope...
Oscar Barenys [Wed, 12 Mar 2025 19:06:58 +0000 (20:06 +0100)]
Update build.yml for Windows Vulkan builder to use Vulkan 1.4.304 SDK for VK_NV_cooperative_matrix2 support (#12301)

3 months agollama.swiftui : fix xcframework dir in README [no ci] (#12353)
Daniel Bevenius [Wed, 12 Mar 2025 12:45:32 +0000 (13:45 +0100)]
llama.swiftui : fix xcframework dir in README [no ci] (#12353)

This commit fixes the path to the xcframework in the README file which I
had forgotten to change after renaming the build directory.

3 months agosycl : variable sg_size support for mmvq kernels (#12336)
Alberto Cabrera Pérez [Wed, 12 Mar 2025 09:57:32 +0000 (09:57 +0000)]
sycl : variable sg_size support for mmvq kernels (#12336)

3 months agoCUDA/HIP: Fix fattn-vec-* when device warp size is not 32 (#12315)
uvos [Wed, 12 Mar 2025 09:14:11 +0000 (10:14 +0100)]
CUDA/HIP: Fix fattn-vec-* when device warp size is not 32 (#12315)

When fattn-wmma was ported over to warp64 various bits that also touch fattn-vec where converted to
selectable warp size, however the fattn-vec kernels dont work with 64 wide warps for now, so we need
to avoid launching them with parameters for warp64

3 months agollama : Add Gemma 3 support (+ experimental vision capability) (#12343)
Xuan-Son Nguyen [Wed, 12 Mar 2025 08:30:24 +0000 (09:30 +0100)]
llama : Add Gemma 3 support (+ experimental vision capability) (#12343)

* llama : Add Gemma 3 text-only support

* fix python coding style

* fix compile on ubuntu

* python: fix style

* fix ubuntu compile

* fix build on ubuntu (again)

* fix ubuntu build, finally

* clip : Experimental support for Gemma 3 vision (#12344)

* clip : Experimental support for Gemma 3 vision

* fix build

* PRId64

3 months agovulkan: fix bug in coopmat1 mul_mat_id (#12316)
Jeff Bolz [Wed, 12 Mar 2025 05:59:19 +0000 (00:59 -0500)]
vulkan: fix bug in coopmat1 mul_mat_id (#12316)

* tests: run mul_mat_id with a larger N

* vulkan: fix bug in coopmat1 mul_mat_id

3 months agoCUDA/HIP: refractor mmqv to unify the calculation of nwarps and rows per block betwee...
uvos [Tue, 11 Mar 2025 19:16:03 +0000 (20:16 +0100)]
CUDA/HIP: refractor mmqv to unify the calculation of nwarps and rows per block between host and device code. (#12177)

refactor mmqv to unify the calculation of nwarps and rows per block between host and device code.

---------

Co-authored-by: Johannes Gäßler <redacted>
3 months agoggml-backend : fix backend search path (#12330)
jklincn [Tue, 11 Mar 2025 13:25:17 +0000 (21:25 +0800)]
ggml-backend : fix backend search path (#12330)

* Fix backend search path

* replace .native() with '/'

* reverted .native()

3 months agometal : Cache the Metal library at the device context level (#12265)
BB-fat [Tue, 11 Mar 2025 11:45:02 +0000 (19:45 +0800)]
metal : Cache the Metal library at the device context level (#12265)

3 months agoclip : bring back GPU support (#12322)
Xuan-Son Nguyen [Tue, 11 Mar 2025 08:20:16 +0000 (09:20 +0100)]
clip : bring back GPU support (#12322)

* clip : bring back GPU support

* use n_gpu_layers param

* fix double free

* ggml_backend_init_by_type

* clean up

3 months agomat vec double buffer (#12188)
Eve [Mon, 10 Mar 2025 19:28:11 +0000 (19:28 +0000)]
mat vec double buffer (#12188)

3 months agomusa: support new arch mp_31 and update doc (#12296)
R0CKSTAR [Mon, 10 Mar 2025 17:18:25 +0000 (01:18 +0800)]
musa: support new arch mp_31 and update doc (#12296)

Signed-off-by: Xiaodong Ye <redacted>
3 months agoopencl: use OpenCL C standard supported by the device (#12221)
Henry Linjamäki [Mon, 10 Mar 2025 16:57:00 +0000 (18:57 +0200)]
opencl: use OpenCL C standard supported by the device (#12221)

This patch nudges the llama.cpp a bit to be supported on PoCL which
doesn't support OpenCL C CL2.0. The issue is solved by querying the
device for the supported OpenCL C versions and using the highest one
available.

3 months agoreadme: added Sidekick to available UIs (#12311)
John Bean [Mon, 10 Mar 2025 14:13:09 +0000 (22:13 +0800)]
readme: added Sidekick to available UIs (#12311)

3 months agotests : fix test-quantize-fns to init the CPU backend (#12306)
Georgi Gerganov [Mon, 10 Mar 2025 12:07:15 +0000 (14:07 +0200)]
tests : fix test-quantize-fns to init the CPU backend (#12306)

ggml-ci

3 months agocommon : refactor '-o' option (#12278)
marcoStocchi [Mon, 10 Mar 2025 11:34:13 +0000 (12:34 +0100)]
common : refactor '-o' option (#12278)

As discussed in PR 'llama-tts : add -o option' (#12042):

* common_params : 'out_file' string is the only output file name parameter left in common_params. It's intended to be used in all example programs implementing an '-o' option.

* cvector-generator, export-lora, imatrix : default output filenames moved from 'common_params' to the 'main()' of each example program.

3 months ago`server`: extract <think> tags from qwq outputs (#12297)
Olivier Chafik [Mon, 10 Mar 2025 10:59:03 +0000 (10:59 +0000)]
`server`: extract <think> tags from qwq outputs (#12297)

* extract <think> tags from qwq outputs

* const for all static regexes in chat.cpp

3 months ago`tool-call`: ensure there's always a non-empty tool call id (#12292)
Olivier Chafik [Mon, 10 Mar 2025 09:45:29 +0000 (09:45 +0000)]
`tool-call`: ensure there's always a non-empty tool call id (#12292)

3 months agoallow missing content in message if tool_calls provided (#12293)
Olivier Chafik [Mon, 10 Mar 2025 09:45:07 +0000 (09:45 +0000)]
allow missing content in message if tool_calls provided (#12293)

3 months ago`sampler`: fixes trigger tokens + lazy grammars (fix typo cast from token to string...
Olivier Chafik [Mon, 10 Mar 2025 09:44:42 +0000 (09:44 +0000)]
`sampler`: fixes trigger tokens + lazy grammars (fix typo cast from token to string) (#12291)

* Fix typo in lazy grammar handling (fixes trigger tokens)

Co-authored-by: Georgi Gerganov <redacted>
---------

Co-authored-by: Georgi Gerganov <redacted>
3 months agollava : fix bug in minicpm-v code (#11513)
tc-mb [Mon, 10 Mar 2025 08:33:24 +0000 (16:33 +0800)]
llava : fix bug in minicpm-v code (#11513)

* fix bug in minicpm-v code

* update readme of minicpm-v

3 months agoserver : add speculative decoding presets for FIM (#12287)
Georgi Gerganov [Sun, 9 Mar 2025 17:08:20 +0000 (19:08 +0200)]
server : add speculative decoding presets for FIM (#12287)

3 months agoauthors : update (#12271)
Georgi Gerganov [Sat, 8 Mar 2025 16:26:00 +0000 (18:26 +0200)]
authors : update (#12271)

3 months agoggml-backend : make path_str compatible with C++20 (#12269)
Jason C.H [Sat, 8 Mar 2025 16:02:39 +0000 (00:02 +0800)]
ggml-backend : make path_str compatible with C++20 (#12269)

3 months agoserver : infill gen ends on new line (#12254)
Georgi Gerganov [Fri, 7 Mar 2025 18:54:30 +0000 (20:54 +0200)]
server : infill gen ends on new line (#12254)

3 months agoggml : skip intermediate .air file when compiling .metallib (#12247)
Daniel Bevenius [Fri, 7 Mar 2025 13:15:27 +0000 (14:15 +0100)]
ggml : skip intermediate .air file when compiling .metallib (#12247)

This commit updates the compilation of default.metallib to skip the
intermediate .air (Apple Intermediate Representation) file.

The motivation for this change is to simplify the custom command a
little and avoid generating and then removing the .air file.

3 months agosync : ggml upstream/0.0.4853
Georgi Gerganov [Fri, 7 Mar 2025 12:00:27 +0000 (14:00 +0200)]
sync : ggml

ggml-ci

3 months agoggml : ggml_compute_forward_concat() for arbitrary tensor type (ggml/1118)
vmobilis [Fri, 7 Mar 2025 08:11:40 +0000 (11:11 +0300)]
ggml : ggml_compute_forward_concat() for arbitrary tensor type (ggml/1118)

* ggml_compute_forward_concat() for arbitrary tensor type

* Check that tensors' type match

* ggml-cpu.c: check type of source tensors

* ggml-cpu.c: move tensor type check to ggml_compute_forward_concat()

* ggml.c: check concatenated tensor type

* Remove tensor type check from ggml_compute_forward_concat() in ggml-cpu.c

..., as it was moved to ggml.c.

3 months agoggml-cpu: faster AVX2 variant for IQ1_M (#12216)
Rémy O [Fri, 7 Mar 2025 11:54:22 +0000 (12:54 +0100)]
ggml-cpu: faster AVX2 variant for IQ1_M (#12216)

3 months agoci : fix save-load test invocations (#12245)
Georgi Gerganov [Fri, 7 Mar 2025 10:19:31 +0000 (12:19 +0200)]
ci : fix save-load test invocations (#12245)

3 months agoserver : Log original chat template parsing error (#12233)
Sigbjørn Skjæret [Fri, 7 Mar 2025 10:15:33 +0000 (11:15 +0100)]
server : Log original chat template parsing error (#12233)

3 months agosync: minja - support QwQ-32B (#12235)
Olivier Chafik [Fri, 7 Mar 2025 09:33:37 +0000 (09:33 +0000)]
sync: minja - support QwQ-32B (#12235)

https://github.com/google/minja/commit/8a76f7815e8a3ae00bd233c2b5a8b7d4e86564ec