LLM-Inference/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/whisper.cpp.git synced 2026-08-12 22:31:22 +04:00
Code Issues Packages Projects Releases Wiki Activity
4,851 Commits 65 Branches 88 Tags
748b84bfa91454caa5793ecb05f10518b1151ef9
Commit Graph
45 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
Neo ZhangandGeorgi Gerganov 67ed569d31 support the missed types in cpy (llama/26005) 
* support the missed types in cpy

* use correct funct

* rm unused code
 2026-08-04 13:37:47 +03:00
Neo ZhangGeorgi GerganovNeo Zhang Jianyu
 		fcb5b86659 sycl : support dev2dev memcpy by DEV2DEV_MEMCPY_FORWARD (llama/26234) 
Co-authored-by: Neo Zhang Jianyu <jianyu.zhang@intel.com>
 2026-08-04 13:37:47 +03:00
Neo ZhangandGeorgi Gerganov 2dba496574 Support q2 mul_mat (llama/26231) 
* support q2_0 in mul_mat

* support more q2_0 case
 2026-08-04 13:37:47 +03:00
Neo ZhangandGeorgi Gerganov 7c04837122 fix unsupported UT cases of CONT & CPY (llama/25231) 
* fix unsupported UT cases of CONT & CPY

* update ops.md

* rm unused head file
 2026-07-10 13:06:42 +03:00
Neo ZhangandGeorgi Gerganov bcbc5bc554 support op col2im_1d (llama/25264) 
* support op col2im_1d

* update ops.md

* rm unused words

* update for bf16

* optimize 1%-11% as the review comments

* fix the format issue

* update as the review comments
 2026-07-10 13:06:42 +03:00
Neo ZhangandGeorgi Gerganov a0b8896847 support OP cross_entropy_loss, cross_entropy_loss_back (llama/25236) 
* support OP cross_entropy_loss, cross_entropy_loss_back

* correct format issue
 2026-07-10 13:06:42 +03:00
Neo ZhangandGeorgi Gerganov 120fc3729d fix unsupport ACC UT cases for noncontiguous (llama/25124) 
* fix unsupport ACC UT cases for noncontiguous

* update ops.md
 2026-07-10 13:06:42 +03:00
Neo ZhangandGeorgi Gerganov bfd86f4304 sycl : enhance argsort to support all UT cases (llama/25125) 2026-07-10 13:06:42 +03:00
Neo ZhangandGeorgi Gerganov f353a27988 sycl : use sycl func to fix AOT double type issue (llama/25081) 2026-07-10 13:06:42 +03:00
Neo ZhangandGeorgi Gerganov 2c6e470ced sycl : rename the env vars from "disable" to "enable" (llama/25042) 2026-07-10 13:06:42 +03:00
Neo ZhangandGeorgi Gerganov a4ac1c0369 sycl : fix failed ut cases of norm (llama/25044) 2026-07-10 13:06:42 +03:00
Neo ZhangandGeorgi Gerganov c4ba5e2e13 sycl : fix the failed UT cases of conv_3d (llama/24900) 2026-06-26 16:03:57 +03:00
Neo ZhangandGeorgi Gerganov 1b3f8434f9 support bf16 on bin_bcast OP and unary OPs (llama/24838) 
* support bf16 on bin_bcast OP and unary OPs

* support the older Intel compiler than 2026.0
 2026-06-26 16:03:57 +03:00
Neo ZhangandGeorgi Gerganov 4c4f6eaca7 rename GGML_SYCL_SUPPORT_LEVEL_ZERO (llama/24719) 
* rename GGML_SYCL_SUPPORT_LEVEL_ZERO to GGML_SYCL_SUPPORT_LEVEL_ZERO_API, and GGML_SYCL_ENABLE_LEVEL_ZERO to  GGML_SYCL_USE_LEVEL_ZERO_API

* fix code format

* fix error when rebase
 2026-06-19 12:53:43 +03:00
Neo ZhangandGeorgi Gerganov 69a9798a82 sycl : support MUL_MAT and OUT_PROD with Q1_0 (llama/24721) 2026-06-19 12:53:43 +03:00
Neo ZhangandGeorgi Gerganov 312033789f support OPs: conv_2d, conv_2d_dw, conv2d_transpose (llama/24600) 
* fix conflict

* fix format issue, rename

* rm debug code

* correct the file name
 2026-06-19 12:53:43 +03:00
Neo ZhangGeorgi GerganovNeo Zhang <NA>
 		6772827b26 add dev2dev memcpy by SYCL API (llama/24476) 
* add dev2dev memcpy by SYCL API

* mv GGML_SYCL_DEV2DEV_MEMCPY to runntime table

* update the detect method for p2p comm

* fix the erro created during fix confilct

---------

Co-authored-by: Neo Zhang <NA>
 2026-06-19 12:53:43 +03:00
Neo ZhangandGeorgi Gerganov 01b1c3ada7 Add conv_3d (llama/24691) 
* add conv_3d

* optimize

* update ops.md

* restore test script

* rm unused code

* rm copyright notes
 2026-06-19 12:53:43 +03:00
Neo ZhangandGeorgi Gerganov da66f048b1 sycl : Enable to support fp16 by OPs: SQR, SQRT, LOG, SIN, COS, CLAMP (llama/24692) 2026-06-19 12:53:43 +03:00
Neo ZhangandGeorgi Gerganov 79f88a1104 Support OP EXPM1, support all UT cases of FLOOR, TRUNC, ROUND (llama/24363) 
* support OP EXPM1, support all UT cases of FLOOR, TRUNC, ROUND

* fix conflict

* rebase, support new UT case of repeat, concat
 2026-06-19 12:53:43 +03:00
Neo ZhangandGeorgi Gerganov e958dcead1 sycl : fix reorder function; add fp32/fp16 in build script (llama/24578) 2026-06-19 12:53:43 +03:00
Neo ZhangandGeorgi Gerganov d20057908a sycl : enhance set_rows to support q1_0, mxfp4, nvfp4 (llama/24564) 2026-06-19 12:53:43 +03:00
Neo ZhangandGeorgi Gerganov 3cb087c42a add to support pool_1d, move pool_1d/2d code to pool.cpp/hpp (llama/24584) 
* add to support pool_1d, move pool_1d/2d code to pool.cpp/hpp

* update ops.md
 2026-06-19 12:53:43 +03:00
Neo ZhangandGeorgi Gerganov ec0c661950 Support Q4_1, Q5_0, Q5_1 in Flash-attention (llama/23812) 
* support Q4_1, Q5_0, Q5_1

* update ut case
 2026-06-08 14:36:36 +03:00
Neo ZhangandGeorgi Gerganov 20323e48c4 Add more types in GET_ROWS OP (llama/23710) 
* add to support Q1_0, NVFP4, IQ2_XXS, IQ2_XS, IQ2_S, IQ3_XXS, IQ1_S, IQ1_M, IQ3_S, IQ4_NL, IQ4_XS, I32, MXFP4, Q2_K, Q3_K, Q5_K, and Q6_K in GET_ROWS OP

* correct the link
 2026-06-08 14:36:36 +03:00
Neo ZhangandGeorgi Gerganov 687fbcb149 sycl : Optimize Q3_K mul_mat by reorder (llama/23725) 2026-06-08 14:36:36 +03:00
Neo ZhangandGeorgi Gerganov 4db2f45075 Add OP im2col_3d (llama/22903) 
* add im2col_3d

* format code

* update the ops.md
 2026-05-14 21:26:48 +03:00
Neo ZhangandGeorgi Gerganov 6296fd5a90 Optimize Q4_0 mul_mat for Arc770, add scripts (llama/22291) 
* opt arc770 for Q4_0

* add for Q4_0

* update the script

* add help script for windows

* update guide

* fix format issue

* convert from dos to unix for format issue

* fix missed -sm parameter
 2026-04-30 11:29:19 +03:00
Neo ZhangandGeorgi Gerganov 42e4a28865 sycl : handle other FA case (llama/21377) 2026-04-30 11:29:01 +03:00
Neo ZhangandGeorgi Gerganov 7f6c0ac20f sycl : fix llama_kv_cache hang when kv_cache is huge: 5GB (llama/21283) 2026-04-30 11:29:00 +03:00
Neo ZhangandGeorgi Gerganov fab70d287e sycl : support nvfp4 type in mul_mat (llama/21227) 2026-04-30 11:28:59 +03:00
Neo ZhangandGeorgi Gerganov 952c66237d sycl : enhance fattn perf (llama/21185) 2026-04-30 11:28:57 +03:00
Neo ZhangandGeorgi Gerganov f2a8e65ea7 sycl : fix wrong variable check by assert (llama/20903) 
* fix wrong variable check by assert

* use GGML api
 2026-03-29 15:04:36 +03:00
Neo ZhangandGeorgi Gerganov 1d0f0285de support bf16 and quantized type (llama/20803) 2026-03-29 15:04:36 +03:00
Neo ZhangandGeorgi Gerganov 6494251197 ehance UPSCALE to support all UT cases (llama/20637) 
* [SYCL] ehance UPSCALE to support more cases

* rm test case result of SYCL1
 2026-03-29 15:04:36 +03:00
Neo ZhangandGeorgi Gerganov c5f9a49b51 add op gated_delta_net (llama/20455) 2026-03-16 13:10:15 +02:00
Neo ZhangandGeorgi Gerganov 286387ef0a fix op rope, add rope_back (llama/20293) 2026-03-16 13:10:15 +02:00
Neo ZhangandGeorgi Gerganov 72c7a2532d fix for failed UT case: ACC, L2_NORM, UPSCALE, fused_glu, unary (llama/20283) 2026-03-16 13:10:15 +02:00
Neo ZhangandGeorgi Gerganov 8a9b0ba1df supprt Flash Attention for fp32/fp16/Q4/Q5/Q8 (llama/20190) 
* support flash-attention for fp32/fp16/Q4/Q5/Q8

* rm warining

* update for JIT
 2026-03-16 13:10:15 +02:00
Neo ZhangGeorgi GerganovNeo Zhang Jianyu
 		64f48603e6 replace the magic nunber 768 by max work group size to support iGPU (llama/19920) 
Co-authored-by: Neo Zhang Jianyu <jianyu.zhang@intel.com>
 2026-02-27 20:57:58 +02:00
Neo ZhangGeorgi GerganovNeo Zhang Jianyu
 		4cac408c60 support permuted, remove check s0/s10 (llama/19889) 
Co-authored-by: Neo Zhang Jianyu <jianyu.zhang@intel.com>
 2026-02-27 20:57:58 +02:00
Neo ZhangandGeorgi Gerganov c4003da2b8 Remove support for Nvidia & AMD GPU, because the oneAPI plugin for Nvidia & AMD GPU is unavailable: download/installation channels are out of work. (llama/19246) 
User can't build up the software for Nvidia & AMD GPU.
rm the oneMath since it is only used in NV and AMD code path.
 2026-02-08 09:29:10 +02:00
Neo ZhangGeorgi GerganovNeo Zhang Jianyu
 		f0e85bb142 sycl: fix norm kernels: l2_norm, group_norm, rms_norm by remove assert to support more cases (llama/19154) 
Co-authored-by: Neo Zhang Jianyu <jianyu.zhang@intel.com>
 2026-01-30 15:56:40 +02:00
Neo ZhangGeorgi GerganovNeo Zhang Jianyu
 		0d9dda5a99 use malloc to support both iGPU and dGPU in same time (llama/18992) 
* use malloc to support both iGPU and dGPU in same time

* support windows

---------

Co-authored-by: Neo Zhang Jianyu <jianyu.zhang@intel.com>
 2026-01-30 15:56:40 +02:00
Neo ZhangGeorgi GerganovNeo Zhang Jianyu
 		a3459484bf sycl : support to malloc memory on device more than 4GB, update the doc and script (llama/17566) 
Co-authored-by: Neo Zhang Jianyu <jianyu.zhang@intel.com>
 2025-12-12 17:53:13 +02:00