mirror/whisper.cpp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
479bd77169
whisper.cpp/ggml/src/ggml-sycl
History
Djip007 e990d1b791 ggml : refactor online repacking (llama/10446) 
* rename ggml-cpu-aarch64.c to .cpp

* reformat extra cpu backend.

- clean Q4_0_N_M and IQ4_0_N_M
  - remove from "file" tensor type
  - allow only with dynamic repack

- extract cpu extra bufts and convert to C++
  - hbm
  - "aarch64"

- more generic use of extra buffer
  - generalise extra_supports_op
  - new API for "cpu-accel":
     - amx
     - aarch64

* clang-format

* Clean Q4_0_N_M ref

Enable restrict on C++

* add op GGML_OP_MUL_MAT_ID for Q4_0_N_M with runtime repack

* added/corrected control on tensor size for Q4 repacking.

* Update ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

* Update ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

* add debug logs on repacks.

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2024-12-18 12:52:16 +02:00
..
dpct SYCL : Move to compile time oneMKL interface backend selection for NVIDIA backend (llama/10584) 2024-12-08 20:14:35 +02:00
backend.hpp Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (llama/10133) 2024-11-15 15:21:04 +02:00
CMakeLists.txt SYCL : Move to compile time oneMKL interface backend selection for NVIDIA backend (llama/10584) 2024-12-08 20:14:35 +02:00
common.cpp Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (llama/10133) 2024-11-15 15:21:04 +02:00
common.hpp Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (llama/10133) 2024-11-15 15:21:04 +02:00
concat.cpp Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (llama/10133) 2024-11-15 15:21:04 +02:00
concat.hpp add concat through dim 1/2 (llama/8483) 2024-08-08 22:48:46 +03:00
conv.cpp add conv support (llama/8688) 2024-08-08 22:48:46 +03:00
conv.hpp add conv support (llama/8688) 2024-08-08 22:48:46 +03:00
convert.cpp Fix SYCL im2col and convert Overflow with Large Dims (llama/9052) 2024-08-28 13:22:20 +03:00
convert.hpp Fix SYCL im2col and convert Overflow with Large Dims (llama/9052) 2024-08-28 13:22:20 +03:00
dequantize.hpp Fixed dequant precision issues in Q4_1 and Q5_1 (llama/9711) 2024-10-05 15:23:51 +03:00
dmmv.cpp Fix DMMV dequantization (llama/9279) 2024-09-24 19:45:08 +03:00
dmmv.hpp whisper : reorganize source code + improve CMake (#2256) 2024-06-26 19:34:09 +03:00
element_wise.cpp Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (llama/10133) 2024-11-15 15:21:04 +02:00
element_wise.hpp Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (llama/10133) 2024-11-15 15:21:04 +02:00
gemm.hpp Add oneDNN primitive support (llama/9091) 2024-08-28 13:22:20 +03:00
ggml-sycl.cpp ggml : refactor online repacking (llama/10446) 2024-12-18 12:52:16 +02:00
im2col.cpp Fix SYCL im2col and convert Overflow with Large Dims (llama/9052) 2024-08-28 13:22:20 +03:00
im2col.hpp Fix SYCL im2col and convert Overflow with Large Dims (llama/9052) 2024-08-28 13:22:20 +03:00
mmq.cpp ggml : reduce hash table reset cost (llama/8698) 2024-08-08 22:48:46 +03:00
mmq.hpp whisper : reorganize source code + improve CMake (#2256) 2024-06-26 19:34:09 +03:00
mmvq.cpp fix mul_mat_vec_q and *_vec_q error (llama/9939) 2024-11-01 10:19:05 +02:00
mmvq.hpp whisper : reorganize source code + improve CMake (#2256) 2024-06-26 19:34:09 +03:00
norm.cpp sycl : Fixes to broken builds and test-backend-ops (llama/10257) 2024-11-15 15:21:04 +02:00
norm.hpp ggml : sync sycl (skip) (#0) 2024-07-08 14:53:55 +03:00
outprod.cpp SYCL : Move to compile time oneMKL interface backend selection for NVIDIA backend (llama/10584) 2024-12-08 20:14:35 +02:00
outprod.hpp Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (llama/10133) 2024-11-15 15:21:04 +02:00
presets.hpp Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (llama/10133) 2024-11-15 15:21:04 +02:00
rope.cpp ggml : move rope type enum to ggml.h (llama/8949) 2024-08-28 13:22:20 +03:00
rope.hpp ggml : sync sycl (skip) (#0) 2024-07-08 14:53:55 +03:00
softmax.cpp fix scratch size of softmax (llama/8642) 2024-08-08 22:48:46 +03:00
softmax.hpp ggml : sync sycl (skip) (#0) 2024-07-08 14:53:55 +03:00
tsembd.cpp Add TIMESTEP_EMBEDDING OP (llama/8707) 2024-08-08 22:48:46 +03:00
tsembd.hpp Add TIMESTEP_EMBEDDING OP (llama/8707) 2024-08-08 22:48:46 +03:00
vecdotq.hpp sycl: Use syclcompat::dp4a (llama/10267) 2024-11-20 21:00:08 +02:00
wkv6.cpp Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (llama/10133) 2024-11-15 15:21:04 +02:00
wkv6.hpp Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (llama/10133) 2024-11-15 15:21:04 +02:00