]> git.djapps.eu Git - pkg/ggml/sources/whisper.cpp/log
pkg/ggml/sources/whisper.cpp
23 months agoAllow all RDNA2 archs to use sdot4 intrinsic (llama/8629)
Jeroen Mostert [Tue, 23 Jul 2024 08:50:40 +0000 (10:50 +0200)]
Allow all RDNA2 archs to use sdot4 intrinsic (llama/8629)

The check gating the use of `__builtin_amdgc_sdot4` specifically checks for gfx1030. This causes a severe perf regression for anything gfx103? that's not gfx1030 and not using `HSA_OVERRIDE_GFX_VERSION` (if you've built ROCm to support it). We already have a generic RDNA2 define, let's use it.

23 months agofix scratch size of softmax (llama/8642)
luoyu-intel [Tue, 23 Jul 2024 07:43:28 +0000 (07:43 +0000)]
fix scratch size of softmax (llama/8642)

23 months agoggml: fix compile error for RISC-V (llama/8623)
Mark Zhuang [Mon, 22 Jul 2024 07:56:45 +0000 (15:56 +0800)]
ggml: fix compile error for RISC-V (llama/8623)

23 months agoCUDA: MMQ code deduplication + iquant support (llama/8495)
Johannes Gäßler [Sat, 20 Jul 2024 20:25:26 +0000 (22:25 +0200)]
CUDA: MMQ code deduplication + iquant support (llama/8495)

* CUDA: MMQ code deduplication + iquant support

* 1 less parallel job for CI build

23 months agogguf : handle null name during init (llama/8587)
Georgi Gerganov [Sat, 20 Jul 2024 14:15:42 +0000 (17:15 +0300)]
gguf : handle null name during init (llama/8587)

23 months agoggml : fix quant dot product with odd number of blocks (llama/8549)
slaren [Fri, 19 Jul 2024 15:17:27 +0000 (17:17 +0200)]
ggml : fix quant dot product with odd number of blocks (llama/8549)

* ggml : fix iq4_nl dot product with odd number of blocks

* ggml : fix odd blocks for ARM_NEON (llama/8556)

* ggml : fix iq4_nl dot product with odd number of blocks

* ggml : fix q4_1

* ggml : fix q5_0

* ggml : fix q5_1

* ggml : fix iq4_nl metal

ggml-ci

* ggml : fix q4_0

* ggml : fix q8_0

ggml-ci

* ggml : remove special Q4_0 code for first 2 blocks

* ggml : fix sumf redefinition

---------

Co-authored-by: slaren <redacted>
---------

Co-authored-by: Georgi Gerganov <redacted>
23 months agoggml : add friendlier error message to fopen errors (llama/8575)
Clint Herron [Fri, 19 Jul 2024 11:05:45 +0000 (07:05 -0400)]
ggml : add friendlier error message to fopen errors (llama/8575)

* Add additional error information when model files fail to load.

* Adding additional error information to most instances of fopen.

23 months agoCUDA: fix partial offloading for ne0 % 256 != 0 (llama/8572)
Johannes Gäßler [Thu, 18 Jul 2024 21:48:47 +0000 (23:48 +0200)]
CUDA: fix partial offloading for ne0 % 256 != 0 (llama/8572)

23 months agocmake : install all ggml public headers (llama/8480)
65a [Thu, 18 Jul 2024 14:47:12 +0000 (07:47 -0700)]
cmake : install all ggml public headers (llama/8480)

Co-authored-by: 65a <redacted>
23 months agoAdd Ascend NPU backend (llama/6035)
hipudding [Wed, 17 Jul 2024 11:23:50 +0000 (19:23 +0800)]
Add Ascend NPU backend (llama/6035)

* [CANN] Add Ascend NPU backend

Ascend is a full-stack AI computing infrastructure for industry
applications and services based on Huawei Ascend processors and
software.

CANN (Compute Architecture of Neural Networks), developped by
Huawei, is a heterogeneous computing architecture for AI.

Co-authored-by: wangshuai09 <redacted>
* delete trailing whitespaces

* Modify the code based on review comment

* Rename LLAMA_CANN to GGML_CANN

* Make ggml-common.h private

* add ggml_cann prefix for acl funcs

* Add logging for CANN backend

* Delete Trailing whitespace

---------

Co-authored-by: wangshuai09 <redacted>
23 months agomake/cmake: add missing force MMQ/cuBLAS for HIP (llama/8515)
Johannes Gäßler [Tue, 16 Jul 2024 19:20:59 +0000 (21:20 +0200)]
make/cmake: add missing force MMQ/cuBLAS for HIP (llama/8515)

23 months agoRefactor lora adapter support (llama/8332)
Xuan Son Nguyen [Mon, 15 Jul 2024 18:50:47 +0000 (20:50 +0200)]
Refactor lora adapter support (llama/8332)

* lora: load to devide buft

* add patch tensor function

* correct tensor patch

* llama_lora_adapter_apply

* correct ggml_backend_tensor_copy

* add llm_build_mm

* fix auto merge

* update based on review comments

* add convert script

* no more transpose A

* add f16 convert

* add metadata check

* add sanity check

* fix ftype

* add requirements

* fix requirements

* fix outfile

* conversion: only allow selected models

* fix types

* cuda : do not use dmmv if the tensor does not have enough cols

* llama : lora fixes

* do not disable mmap with lora

Co-authored-by: slaren <redacted>
* llm_build_lora_mm_id

* convert_lora : MoE LoRA conversion support

* convert_lora : prefer safetensors, similarly to convert_hf

* convert_hf : simplify modify_tensors for InternLM2

* convert_lora : lazy conversion

* llama : load and use alpha from LoRA adapters

* llama : use llm_build_lora_mm in most model graphs

* auto scale

* Revert "auto scale"

This reverts commit 42415a4874e0f963e4aca6796ea5dfb97cd17464.

* remove redundant params

* Apply suggestions from code review

Co-authored-by: slaren <redacted>
* change kv metadata

* move add_type to __init__

* convert_hf : move add_type to main()

* convert_lora : use the GGUFWriter from Model instead of overwriting it

---------

Co-authored-by: slaren <redacted>
Co-authored-by: Francis Couture-Harpin <redacted>
23 months agoadd concat through dim 1/2 (llama/8483)
Meng, Hengyu [Mon, 15 Jul 2024 11:32:15 +0000 (19:32 +0800)]
add concat through dim 1/2 (llama/8483)

* add concat through dim 1/2

23 months agoVulkan MMQ Fix (llama/8479)
0cc4m [Mon, 15 Jul 2024 07:38:52 +0000 (09:38 +0200)]
Vulkan MMQ Fix (llama/8479)

* Fix incoherence by adding missing LOAD_VEC_A parameter

* Fix Vulkan op result checker build error

23 months agovulkan : cmake integration (llama/8119)
bandoti [Sat, 13 Jul 2024 16:12:39 +0000 (13:12 -0300)]
vulkan : cmake integration (llama/8119)

* Add Vulkan to CMake pkg

* Add Sycl to CMake pkg

* Add OpenMP to CMake pkg

* Split generated shader file into separate translation unit

* Add CMake target for Vulkan shaders

* Update README.md

* Add make target for Vulkan shaders

* Use pkg-config to locate vulkan library

* Add vulkan SDK dep to ubuntu-22-cmake-vulkan workflow

* Clean up tabs

* Move sudo to apt-key invocation

* Forward GGML_EXTRA_LIBS to CMake config pkg

* Update vulkan obj file paths

* Add shaderc to nix pkg

* Add python3 to Vulkan nix build

* Link against ggml in cmake pkg

* Remove Python dependency from Vulkan build

* code review changes

* Remove trailing newline

* Add cflags from pkg-config to fix w64devkit build

* Update README.md

* Remove trailing whitespace

* Update README.md

* Remove trailing whitespace

* Fix doc heading

* Make glslc required Vulkan component

* remove clblast from nix pkg

23 months agometal : template-ify some of the kernels (llama/8447)
Georgi Gerganov [Sat, 13 Jul 2024 15:32:33 +0000 (18:32 +0300)]
metal : template-ify some of the kernels (llama/8447)

ggml-ci

23 months agoggml : minor naming changes (llama/8433)
Georgi Gerganov [Fri, 12 Jul 2024 07:46:02 +0000 (10:46 +0300)]
ggml : minor naming changes (llama/8433)

* ggml : minor naming changes

ggml-ci

* ggml : use PRId64 [no ci]

* ggml : revert FA K/Q names

23 months agofix the mul_mat_id ut issues (llama/8427)
Chen Xi [Fri, 12 Jul 2024 00:52:04 +0000 (00:52 +0000)]
fix the mul_mat_id ut issues (llama/8427)

* fix part of mul_mat_id

* skip the bfloat 16 sycl ut

Signed-off-by: Chen Xi <redacted>
---------

Signed-off-by: Chen Xi <redacted>
Co-authored-by: Meng, Hengyu <redacted>
Co-authored-by: Chen Xi <redacted>
23 months agoggml : add NVPL BLAS support (ggml/8329) (llama/8425)
Nicholai Tukanov [Thu, 11 Jul 2024 16:49:15 +0000 (11:49 -0500)]
ggml : add NVPL BLAS support (ggml/8329) (llama/8425)

* ggml : add NVPL BLAS support

* ggml : replace `<BLASLIB>_ENABLE_CBLAS` with `GGML_BLAS_USE_<BLASLIB>`

---------

Co-authored-by: ntukanov <redacted>
23 months agocuda : suppress 'noreturn' warn in no_device_code (llama/8414)
Daniel Bevenius [Thu, 11 Jul 2024 15:53:42 +0000 (17:53 +0200)]
cuda : suppress 'noreturn' warn in no_device_code (llama/8414)

* cuda : suppress 'noreturn' warn in no_device_code

This commit adds a while(true) loop to the no_device_code function in
common.cuh. This is done to suppress the warning:

```console
/src/ggml-cuda/template-instances/../common.cuh:346:1: warning:
function declared 'noreturn' should not return [-Winvalid-noreturn]
  346 | }
      | ^
```

The motivation for this is to reduce the number of warnings when
compilng with GGML_HIPBLAS=ON.

Signed-off-by: Daniel Bevenius <redacted>
* squash! cuda : suppress 'noreturn' warn in no_device_code

Update __trap macro instead of using a while loop to suppress the
warning.

Signed-off-by: Daniel Bevenius <redacted>
---------

Signed-off-by: Daniel Bevenius <redacted>
23 months agoCUDA: optimize and refactor MMQ (llama/8416)
Johannes Gäßler [Thu, 11 Jul 2024 14:47:47 +0000 (16:47 +0200)]
CUDA: optimize and refactor MMQ (llama/8416)

* CUDA: optimize and refactor MMQ

* explicit q8_1 memory layouts, add documentation

23 months agoUse multi_ptr to clean up deprecated warnings (llama/8256)
AidanBeltonS [Wed, 10 Jul 2024 15:10:49 +0000 (16:10 +0100)]
Use multi_ptr to clean up deprecated warnings (llama/8256)

23 months agoggml : move sgemm sources to llamafile subfolder (llama/8394)
Georgi Gerganov [Wed, 10 Jul 2024 12:23:29 +0000 (15:23 +0300)]
ggml : move sgemm sources to llamafile subfolder (llama/8394)

ggml-ci

23 months agoggml : add AArch64 optimized GEMV and GEMM Q4 kernels (llama/5780)
Dibakar Gope [Wed, 10 Jul 2024 12:14:51 +0000 (07:14 -0500)]
ggml : add AArch64 optimized GEMV and GEMM Q4 kernels (llama/5780)

* Arm AArch64: optimized GEMV and GEMM kernels for q4_0_q8_0, and q8_0_q8_0 quantization

* Arm AArch64: add optimized GEMV and GEMM asm kernels for q4_0_q8_0 quantization and refactor code to address llama.cpp pr#5780 suggestions

* Arm AArch64: add optimized GEMV and GEMM asm kernels for q4_0_q8_0 quantization and refactor code to address llama.cpp pr#5780 suggestions

* Arm AArch64: add optimized GEMV and GEMM asm kernels for q4_0_q8_0 quantization and refactor code to address llama.cpp pr#5780 suggestions

* Arm AArch64: add optimized GEMV and GEMM asm kernels for q4_0_q8_0 quantization and refactor code to address llama.cpp pr#5780 suggestions

* Arm AArch64: add copyright claim only to ggml-aarch64.cpp and ggml-aarch64.h files

* Arm AArch64: minor code refactoring for rebase

* Arm AArch64: minor code refactoring for resolving a build issue with cmake

* Arm AArch64: minor code refactoring to split the Q4_0_AARC64 type into three separate types: Q4_0_4_4, Q4_0_4_8, and Q4_0_8_8

* Arm AArch64: minor code change for resolving a build issue with server-windows

* retrigger checks

* Arm AArch64: minor code changes for rebase

* Arm AArch64: minor changes to skip the pr#7433 vec_dot code for arm cpus with SVE VL not equal to 256 bits

* Arm AArch64: remove stale LLAMA_QKK_64 from CMakeLists.txt and delete build.zig

* Arm AArch64: add reference scalar gemm and gemv, and avoid dynamic memory allocations during quantization for Q4_0_4_4, Q4_0_4_8, and Q4_0_8_8

* Arm AArch64: add multithreaded quantization support for the new types: Q4_0_4_4, Q4_0_4_8, and Q4_0_8_8

* Arm AArch64: minor code refactoring

* Arm AArch64: simplify logic for calling gemm and gemv functions in ggml_compute_forward_mul_mat

* Arm AArch64: minimize changes in ggml_compute_forward_mul_mat

* Arm AArch64: minor code refactoring, and add reference scalar code to quantize routines for new quant types

* Arm AArch64: minor code refactoring

* Arm AArch64: minor code refactoring

* Arm AArch64: minor code refactoring

* rebase on the latest master commit 3fd62a6 and adapt to the new directory structure

* Arm AArch64: remove a redundant comment

* Arm AArch64: add pragma in ggml-aarch64.c to turn -Woverlength-strings warning off

* Arm AArch64: use __aarch64__ check to guard 64-bit neon kernels

* Arm AArch64: update docs/build.md README to include compile time flags for buiilding the Q4_0_4_4 quant type

23 months agosycl : Reenabled mmvq path for the SYCL Nvidia Backend (llama/8372)
Alberto Cabrera Pérez [Tue, 9 Jul 2024 14:03:15 +0000 (15:03 +0100)]
sycl : Reenabled mmvq path for the SYCL Nvidia Backend (llama/8372)

* SYCL : Reenabled mmvq path for the SYCL Nvidia Backend

* Reduced verbosity of comment

23 months agosycl : fix powf call in device code (llama/8368)
Alberto Cabrera Pérez [Mon, 8 Jul 2024 13:22:41 +0000 (14:22 +0100)]
sycl : fix powf call in device code (llama/8368)

23 months agoggml : loop tiling optimizations for scalar path (ggml/898)
Mahesh Madhav [Thu, 25 Jul 2024 07:54:08 +0000 (00:54 -0700)]
ggml : loop tiling optimizations for scalar path (ggml/898)

Apply a loop tiling technique to the generic path, which provides
performance upside for ISAs with enough registers to take advantage
of it. Also helps the compiler optimize this path.

23 months agoggml: add support for float16 input tensors in pooling operations (ggml/895)
Ivan Filipov [Mon, 22 Jul 2024 11:32:02 +0000 (14:32 +0300)]
ggml: add support for float16 input tensors in pooling operations (ggml/895)

* Add support for float16 tensors in 1d pooling operations

* Add support for float16 input tensors in 2d pooling operations

* code cleanup

remove unnecessary casting during srow ptr initialization

---------

Co-authored-by: vanaka11 <redacted>
23 months agovulkan : initialize vk_buffer_struct members to VK_NULL_HANDLE (ggml/893)
Tony Wasserka [Sat, 20 Jul 2024 18:49:44 +0000 (20:49 +0200)]
vulkan : initialize vk_buffer_struct members to VK_NULL_HANDLE (ggml/893)

This prevents invalid frees when destroying a partially initialized
vk_buffer_struct. For example, this could happen in ggml_vk_create_buffer
when running out of device memory.

Co-authored-by: Tony Wasserka <redacted>
23 months agocmake : only enable GGML_NATIVE and x86 flags if not crosscompiling (ggml/885)
Borislav Stanimirov [Fri, 12 Jul 2024 14:24:20 +0000 (17:24 +0300)]
cmake : only enable GGML_NATIVE and x86 flags if not crosscompiling (ggml/885)

23 months agoscripts : sync new files (#0)
Georgi Gerganov [Thu, 8 Aug 2024 11:00:51 +0000 (14:00 +0300)]
scripts : sync new files (#0)

23 months agocmake : fix compile in xcode (#2311)
Daven Sanassy [Mon, 5 Aug 2024 06:48:26 +0000 (07:48 +0100)]
cmake : fix compile in xcode (#2311)

23 months agowhisper : handle empty mel (#2324)
Georgi Gerganov [Sat, 27 Jul 2024 17:35:04 +0000 (20:35 +0300)]
whisper : handle empty mel (#2324)

2 years agowhisper : use vulkan as gpu backend when available (#2302)
Matt Stephenson [Tue, 16 Jul 2024 07:21:09 +0000 (03:21 -0400)]
whisper : use vulkan as gpu backend when available (#2302)

* ggml: use vulkan as gpu backend when available

Signed-off-by: Matt Stephenson <redacted>
* whisper: enable using vk as default buffer type

Signed-off-by: Matt Stephenson <redacted>
---------

Signed-off-by: Matt Stephenson <redacted>
2 years agowhisper : fix DTW assert (#2299)
arizhih [Mon, 15 Jul 2024 12:50:36 +0000 (14:50 +0200)]
whisper : fix DTW assert (#2299)

2 years agocmake : use WHISPER_EXTRA_FLAGS (#2294)
Georgi Gerganov [Tue, 9 Jul 2024 15:54:18 +0000 (18:54 +0300)]
cmake : use WHISPER_EXTRA_FLAGS (#2294)

2 years agocmake : allow external ggml
Borislav Stanimirov [Mon, 8 Jul 2024 14:08:55 +0000 (17:08 +0300)]
cmake : allow external ggml

2 years agocmake : try to fix openvino build (#2281)
Georgi Gerganov [Mon, 8 Jul 2024 12:36:51 +0000 (15:36 +0300)]
cmake : try to fix openvino build (#2281)

2 years agocmake : remove install of llama convert script [no ci] (#2266)
Georgi Gerganov [Mon, 8 Jul 2024 11:21:04 +0000 (14:21 +0300)]
cmake : remove install of llama convert script [no ci] (#2266)

2 years agomake : remove llama prints [no ci] (#2265)
Georgi Gerganov [Mon, 8 Jul 2024 11:19:36 +0000 (14:19 +0300)]
make : remove llama prints [no ci] (#2265)

2 years agotalk-llama : sync llama.cpp
Georgi Gerganov [Mon, 8 Jul 2024 11:14:17 +0000 (14:14 +0300)]
talk-llama : sync llama.cpp

2 years agoexamples : fix compile warnings [no ci] (#0)
Georgi Gerganov [Mon, 8 Jul 2024 11:09:09 +0000 (14:09 +0300)]
examples : fix compile warnings [no ci] (#0)

2 years agosync : ggml
Georgi Gerganov [Mon, 8 Jul 2024 10:50:28 +0000 (13:50 +0300)]
sync : ggml

2 years agoggml : sync sycl (skip) (#0)
Georgi Gerganov [Mon, 8 Jul 2024 10:50:14 +0000 (13:50 +0300)]
ggml : sync sycl (skip) (#0)

2 years agoscripts : fix sync scripts
Georgi Gerganov [Mon, 8 Jul 2024 10:48:14 +0000 (13:48 +0300)]
scripts : fix sync scripts

2 years agoggml : remove unnecessary UNUSED macro call (ggml/880)
Daniel Bevenius [Mon, 8 Jul 2024 10:03:42 +0000 (12:03 +0200)]
ggml : remove unnecessary UNUSED macro call (ggml/880)

This commit removes an UNUSED macro call that is not needed as the
variable n0 is used in the code and will not produce a warning.

Signed-off-by: Daniel Bevenius <redacted>
2 years agocmake : add GGML_BUILD and GGML_SHARED macro definitions (llama/8281)
Natsu [Fri, 5 Jul 2024 14:29:35 +0000 (22:29 +0800)]
cmake : add GGML_BUILD and GGML_SHARED macro definitions (llama/8281)

2 years agoEnabled more data types for oneMKL gemm_batch (llama/8236)
Ouadie EL FAROUKI [Fri, 5 Jul 2024 12:23:25 +0000 (13:23 +0100)]
Enabled more data types for oneMKL gemm_batch (llama/8236)

2 years agoCUDA: MMQ support for iq4_nl, iq4_xs (llama/8278)
Johannes Gäßler [Fri, 5 Jul 2024 07:06:31 +0000 (09:06 +0200)]
CUDA: MMQ support for iq4_nl, iq4_xs (llama/8278)

2 years agoCUDA: revert part of the RDNA1 optimizations (llama/8309)
Daniele [Fri, 5 Jul 2024 07:06:09 +0000 (07:06 +0000)]
CUDA: revert part of the RDNA1 optimizations (llama/8309)

The change on the launch_bounds was causing a small performance drop in perplexity of 25 t/s

2 years agoCUDA: fix MMQ stream-k rounding if ne00 % 128 != 0 (llama/8311)
Johannes Gäßler [Fri, 5 Jul 2024 07:05:34 +0000 (09:05 +0200)]
CUDA: fix MMQ stream-k rounding if ne00 % 128 != 0 (llama/8311)

2 years agoFix WARP_SIZE=16 bug of Intel GPU (llama/8266)
luoyu-intel [Fri, 5 Jul 2024 05:06:13 +0000 (05:06 +0000)]
Fix WARP_SIZE=16 bug of Intel GPU (llama/8266)

* fix group_norm ut

* split softmax

* fix softmax

* add concat support condition

* revert debug code

* move QK_WARP_SIZE to presets.hpp

2 years agorm get_work_group_size() by local cache for performance (llama/8286)
Neo Zhang Jianyu [Fri, 5 Jul 2024 02:32:29 +0000 (10:32 +0800)]
rm get_work_group_size() by local cache for performance (llama/8286)

Co-authored-by: arthw <redacted>
2 years agoDefine and optimize RDNA1 (llama/8085)
Daniele [Wed, 3 Jul 2024 23:02:58 +0000 (23:02 +0000)]
Define and optimize RDNA1 (llama/8085)

2 years agofix typo (llama/8267)
Judd [Wed, 3 Jul 2024 12:40:16 +0000 (20:40 +0800)]
fix typo (llama/8267)

Co-authored-by: Judd <redacted>
2 years agoRemoves multiple newlines at the end of files that is breaking the editorconfig step...
Clint Herron [Tue, 2 Jul 2024 16:18:10 +0000 (12:18 -0400)]
Removes multiple newlines at the end of files that is breaking the editorconfig step of CI. (llama/8258)

2 years agocuda : update supports_op for matrix multiplication (llama/8245)
slaren [Tue, 2 Jul 2024 06:39:38 +0000 (08:39 +0200)]
cuda : update supports_op for matrix multiplication (llama/8245)

2 years agoFix win build conflict of math library (llama/8230)
luoyu-intel [Tue, 2 Jul 2024 04:50:07 +0000 (04:50 +0000)]
Fix win build conflict of math library (llama/8230)

* fix win build conflict of math library

* fix the condition: !(win32 & SYCL)

* revert warp_size=16

2 years agoFix the sub group size of Intel (llama/8106)
luoyu-intel [Tue, 2 Jul 2024 02:16:00 +0000 (02:16 +0000)]
Fix the sub group size of Intel (llama/8106)

* use warp_size macro for all sycl kernels

* fix mask of permute_sub_group_by_xor

* fix rms_norm with correct warp number

* fix rms_norm_f32/group_norm_f32

* move norm to norm.cpp file

* fix quantize bug

* fix mmvq's batch size

2 years agoCUDA: refactor and optimize IQ MMVQ (llama/8215)
Johannes Gäßler [Mon, 1 Jul 2024 18:39:06 +0000 (20:39 +0200)]
CUDA: refactor and optimize IQ MMVQ (llama/8215)

* CUDA: refactor and optimize IQ MMVQ

* uint -> uint32_t

* __dp4a -> ggml_cuda_dp4a

* remove MIN_CC_DP4A checks

* change default

* try CI fix

2 years agoUpdate SYCL-Rope op and Refactor (llama/8157)
zhentaoyu [Mon, 1 Jul 2024 11:39:06 +0000 (19:39 +0800)]
Update SYCL-Rope op and Refactor (llama/8157)

* align with rope.cu and move sycl-op to a single file

2 years agoCUDA: fix MMQ stream-k for --split-mode row (llama/8167)
Johannes Gäßler [Thu, 27 Jun 2024 14:26:05 +0000 (16:26 +0200)]
CUDA: fix MMQ stream-k for --split-mode row (llama/8167)

2 years agofeat: cuda implementation for `ggml_conv_transpose_1d` (ggml/854)
John Balis [Tue, 2 Jul 2024 16:09:52 +0000 (11:09 -0500)]
feat: cuda implementation for `ggml_conv_transpose_1d` (ggml/854)

* conv transpose 1d passing test for 1d input and kernel

* working for different input and output channel counts, added test for variable stride

* initial draft appears to work with stride other than 1

* working with all old and new conv1d  tests

* added a test for large tensors

* removed use cuda hardcoding

* restored test-conv-transpose.c

* removed unused arugments, and fixed bug where test failure would cause subsequent tests to fail

* fixed accumulator bug

* added test to test-backend-ops

* fixed mistake

* addressed review

* fixed includes

* removed blank lines

* style and warning fixes

* return failure when test fails

* fix supports_op

---------

Co-authored-by: slaren <redacted>
2 years agoci : disable java build
Georgi Gerganov [Mon, 8 Jul 2024 11:26:59 +0000 (14:26 +0300)]
ci : disable java build

2 years agoserver : add inference path to make OAI API compatible (#2270)
Emmanuel Schmidbauer [Mon, 8 Jul 2024 11:24:58 +0000 (07:24 -0400)]
server : add inference path to make OAI API compatible (#2270)

2 years agosync : ggml + fix sync script
Georgi Gerganov [Wed, 26 Jun 2024 20:20:19 +0000 (23:20 +0300)]
sync : ggml + fix sync script

2 years agomake : disable CUDA graphs
Georgi Gerganov [Wed, 26 Jun 2024 20:20:13 +0000 (23:20 +0300)]
make : disable CUDA graphs

2 years agoggml : add GGML_CUDA_USE_GRAPHS option, restore GGML_CUDA_FORCE_CUBLAS (cmake) (llama...
slaren [Wed, 26 Jun 2024 19:34:14 +0000 (21:34 +0200)]
ggml : add GGML_CUDA_USE_GRAPHS option, restore GGML_CUDA_FORCE_CUBLAS (cmake) (llama/8140)

2 years agomake : disable CUDA mel build
Georgi Gerganov [Wed, 26 Jun 2024 19:25:25 +0000 (22:25 +0300)]
make : disable CUDA mel build

2 years agocmake : minor fixes
Georgi Gerganov [Wed, 26 Jun 2024 18:42:39 +0000 (21:42 +0300)]
cmake : minor fixes

2 years agomake : fix missing -O3
Georgi Gerganov [Wed, 26 Jun 2024 18:20:45 +0000 (21:20 +0300)]
make : fix missing -O3

same as https://github.com/ggerganov/llama.cpp/pull/8143

2 years agowhisper : disable CUDA mel + fix FFMPEG
Georgi Gerganov [Wed, 26 Jun 2024 17:11:38 +0000 (20:11 +0300)]
whisper : disable CUDA mel + fix FFMPEG

2 years agosync : ggml
Georgi Gerganov [Wed, 26 Jun 2024 16:40:23 +0000 (19:40 +0300)]
sync : ggml

2 years agowhisper : reorganize source code + improve CMake (#2256)
Georgi Gerganov [Wed, 26 Jun 2024 16:34:09 +0000 (19:34 +0300)]
whisper : reorganize source code + improve CMake (#2256)

* scripts : update sync [no ci]

* files : reorganize [no ci]

* sync : llama.cpp

* cmake : link math library

* cmake : build normal ggml library

* files : move headers to include

* objc : fix path to ggml-metal.h

* ci : fix WHISPER_CUDA -> GGML_CUDA

* scripts : sync LICENSE [no ci]

2 years agowhisper : optimize fft() function (#2242)
mky_coder [Tue, 18 Jun 2024 15:10:33 +0000 (23:10 +0800)]
whisper : optimize fft() function (#2242)

Co-authored-by: Mike Fan <redacted>
2 years agotalk-llama : sync llama.cpp
Georgi Gerganov [Tue, 18 Jun 2024 06:45:37 +0000 (09:45 +0300)]
talk-llama : sync llama.cpp

2 years agowhisper : use ggml_backend_sched (#2239)
Georgi Gerganov [Tue, 18 Jun 2024 06:37:20 +0000 (09:37 +0300)]
whisper : use ggml_backend_sched (#2239)

* whisper : use ggml_backend_sched (wip)

* use sched in whisper_allocr

* whisper : single backend in whisper_context

* whisper : remove whisper_state->backends_used

* whisper : remove whisper_context->backend

* whisper : reset scheduler after init

* whisper : fix external encoder (e.g. CoreML)

* whisper : cleanup

* whisper : handle null GPU buffer types + fix sycl

---------

Co-authored-by: slaren <redacted>
2 years agofix : remove extra files
Georgi Gerganov [Sun, 16 Jun 2024 16:23:55 +0000 (19:23 +0300)]
fix : remove extra files

2 years agoscripts : sync ggml-blas
Georgi Gerganov [Sun, 16 Jun 2024 16:23:32 +0000 (19:23 +0300)]
scripts : sync ggml-blas

2 years agobuild : update make / cmake
Georgi Gerganov [Sun, 16 Jun 2024 16:10:20 +0000 (19:10 +0300)]
build : update make / cmake

2 years agosync : ggml
Georgi Gerganov [Sun, 16 Jun 2024 15:40:07 +0000 (18:40 +0300)]
sync : ggml

2 years agomove BLAS to a separate backend (cont) (llama/6210)
slaren [Sun, 16 Jun 2024 10:57:37 +0000 (13:57 +0300)]
move BLAS to a separate backend (cont) (llama/6210)

ggml-ci

2 years agoVulkan Shader Refactor, Memory Debugging Option (llama/7947)
0cc4m [Sun, 16 Jun 2024 05:17:31 +0000 (07:17 +0200)]
Vulkan Shader Refactor, Memory Debugging Option (llama/7947)

* Refactor shaders, extract GLSL code from ggml_vk_generate_shaders.py into vulkan-shaders directory

* Improve debug log code

* Add memory debug output option

* Fix flake8

* Fix unnecessary high llama-3 VRAM use

2 years agoscripts : stop sync whisper example from ggml
Georgi Gerganov [Sun, 16 Jun 2024 15:38:46 +0000 (18:38 +0300)]
scripts : stop sync whisper example from ggml

2 years agocmake : fix sycl build (#0)
Georgi Gerganov [Sun, 16 Jun 2024 14:57:35 +0000 (17:57 +0300)]
cmake : fix sycl build (#0)

2 years agoggml : remove OpenCL (#0)
Georgi Gerganov [Sun, 16 Jun 2024 10:46:12 +0000 (13:46 +0300)]
ggml : remove OpenCL (#0)

2 years agosycl : sync (#0)
Georgi Gerganov [Sun, 16 Jun 2024 10:24:17 +0000 (13:24 +0300)]
sycl : sync (#0)

2 years agocuda : enable CUDA graphs (#0)
Georgi Gerganov [Sun, 16 Jun 2024 10:20:19 +0000 (13:20 +0300)]
cuda : enable CUDA graphs (#0)

2 years agotalk-llama : sync llama.cpp
Georgi Gerganov [Sun, 16 Jun 2024 10:10:54 +0000 (13:10 +0300)]
talk-llama : sync llama.cpp

2 years agocmake : fix CUDA build (#0)
Georgi Gerganov [Sun, 16 Jun 2024 10:07:43 +0000 (13:07 +0300)]
cmake : fix CUDA build (#0)

2 years agosync : ggml
Georgi Gerganov [Sun, 16 Jun 2024 09:43:14 +0000 (12:43 +0300)]
sync : ggml

ggml-ci

2 years agoggml : fix and optimize ppc64le (ggml/849)
Hong Bo PENG [Sun, 16 Jun 2024 08:53:11 +0000 (16:53 +0800)]
ggml : fix and optimize ppc64le (ggml/849)

* fix compile issues introduced by loongarch_asx

* restore quant changes to merge

* fix compile issues introduced by loongarch_asx

* further optimize by using vec_msum & vec_sum4s on ppc64le

2 years agoggml : remove duplicate include of ggml-common.h (ggml/853)
Daniel Bevenius [Sun, 16 Jun 2024 08:51:18 +0000 (10:51 +0200)]
ggml : remove duplicate include of ggml-common.h (ggml/853)

Signed-off-by: Daniel Bevenius <redacted>
2 years agoremove global variables (llama/7710)
Meng, Hengyu [Sat, 15 Jun 2024 06:05:10 +0000 (14:05 +0800)]
remove global variables (llama/7710)

* separate DPCT helpers outside

* replace global variables with context

* remove useless extra

* update mul_mat condition

* remove duplicate buft initialization

* remove duplicate extra and global work group size

* remove useless backend check

* remove duplicated extras

* use macro for group_size and remove cuda-related

2 years agoCUDA: faster q2_K, q3_K MMQ + int8 tensor cores (llama/7921)
Johannes Gäßler [Fri, 14 Jun 2024 16:41:49 +0000 (18:41 +0200)]
CUDA: faster q2_K, q3_K MMQ + int8 tensor cores (llama/7921)

* CUDA: faster q2_K, q3_K MMQ + int8 tensor cores

* try CI fix

* try CI fix

* try CI fix

* fix data race

* rever q2_K precision related changes

2 years agometal : utilize max shared memory for mul_mat_id (llama/7935)
Georgi Gerganov [Fri, 14 Jun 2024 14:14:09 +0000 (17:14 +0300)]
metal : utilize max shared memory for mul_mat_id (llama/7935)

2 years agorpc : fix ggml_backend_rpc_supports_buft() (llama/7918)
Radoslav Gerganov [Thu, 13 Jun 2024 12:18:44 +0000 (15:18 +0300)]
rpc : fix ggml_backend_rpc_supports_buft() (llama/7918)

2 years agomove BLAS to a separate backend (llama/6210)
slaren [Thu, 13 Jun 2024 01:11:35 +0000 (03:11 +0200)]
move BLAS to a separate backend (llama/6210)

* move BLAS to a separate backend

* rename GGML_USE_OPENBLAS to GGML_USE_BLAS

* alloc : reuse same buffer when the same buffer type if used multiple times

* set number of threads automatically for openblas and blis

* sched : print assignments when GGML_SCHED_DEBUG env variable is set

* sched : allow ops with weights on an incompatible buffer type

This will cause the weight to be copied to a backend that supports the
op, which is very costly. The weight should have been stored in a buffer
of a backend that can run the op, but llama.cpp cannot do this
automatically at the moment.

---------

Co-authored-by: Georgi Gerganov <redacted>
2 years agoCUDA: fix broken oob check for FA vec f32 kernel (llama/7904)
Johannes Gäßler [Wed, 12 Jun 2024 15:41:51 +0000 (17:41 +0200)]
CUDA: fix broken oob check for FA vec f32 kernel (llama/7904)

2 years agotests : add non-cont unary tests (llama/7857)
Georgi Gerganov [Wed, 12 Jun 2024 13:00:22 +0000 (16:00 +0300)]
tests : add non-cont unary tests (llama/7857)

* tests : add non-cont unary tests

* ggml : update unary asserts and "supports_op"

ggml-ci