happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
3,151 Commits 518 Branches 5,144 Tags 1 GiB
Commit Graph

3151 Commits

Author SHA1 Message Date
Aidan a235b7c532 Vectorize q load 2024-06-17 10:30:40 +01:00
Aidan 604ef6bf15 Store scales in local mem 2024-06-17 10:26:18 +01:00
Aidan cb3fb42046 Single load for half2 2024-06-17 10:21:16 +01:00
Aidan 4a481556e6 Remove double lines 2024-06-17 10:16:10 +01:00
Joe Todd ff076b8873
Merge pull request #7920 from ggerganov/codeplay/revert-host-alloc 
Revert "use the correct SYCL context for host USM allocations"
 2024-06-14 15:10:33 +01:00
Joe Todd b2c8c831c9
Merge pull request #7919 from ggerganov/codeplay/unify-rope-sycl 
sycl-exp : unify rope neox/norm
 2024-06-14 15:08:23 +01:00
Joe Todd ded54b5d9b Replace powf with sycl::pow in ggml-sycl.cpp 
Signed-off-by: Joe Todd <joe.todd@codeplay.com>
 2024-06-14 13:14:33 +01:00
Joe Todd 18133cab40 Revert "use the correct SYCL context for host USM allocations" 
Manually reverting:
https://github.com/ggerganov/llama.cpp/pull/7858

Signed-off-by: Joe Todd <joe.todd@codeplay.com>
 2024-06-13 12:08:27 +01:00
Joe Todd abd7c7b8c2 Formatting 
Signed-off-by: Joe Todd <joe.todd@codeplay.com>
 2024-06-13 10:36:05 +01:00
Joe Todd 0c0f3f0000 [SYCL] Update unsupported ops 
Rope is only supported for contiguous input data.

Concat currently only supports dim=2

Signed-off-by: Joe Todd <joe.todd@codeplay.com>
 2024-06-13 10:33:34 +01:00
Joe Todd 9b81b57239 [SYCL] unify rope norm/neox 
As per: https://github.com/ggerganov/llama.cpp/pull/7634

Signed-off-by: Joe Todd <joe.todd@codeplay.com>
 2024-06-13 10:30:43 +01:00
Georgi Gerganov a9cae48003
tests : add non-cont unary tests (#7857) 
* tests : add non-cont unary tests

* ggml : update unary asserts and "supports_op"

ggml-ci
 2024-06-12 16:00:22 +03:00
Georgi Gerganov bfaa676b08
ggml : improve ggml_is_contiguous logic (#7856) 
* ggml : improve ggml_is_contiguous logic

ggml-ci

* ggml : support more contiguous cases

ggml-ci
 2024-06-12 15:24:20 +03:00
Georgi Gerganov 704a35b183
server : restore numeric prompts (#7883) 2024-06-12 14:42:29 +03:00
Meng, Hengyu dcf752707d
update intel docker oneapi-basekit to 2024.1.1-devel-ubuntu22.04 (#7894) 
In addition this reverts a workaround we had to do to workaround the upstream issue with expired intel GPG package keys in 2024.0.1-devel-ubuntu22.04
 2024-06-12 19:05:35 +10:00
Patrice Ferlet f2b5764beb
Fix a typo and add Fedora 40 pacakge to install for Vulkan (#7794) [no ci] 
Fix "appropiate" to "appropriate" and add Fedora 40 packages to install to compile with Vulkan support
 2024-06-12 11:18:16 +10:00
k.h.lai 73bac2b11d
vulkan: select only one device for single gpu with multiple drivers (#7582) 2024-06-11 21:26:05 +02:00
0cc4m ef52d1d16a
Update Vulkan RoPE implementation (#7818) 
* Update Vulkan RoPE implementation

* Return nullptr on alloc_buffer when allocation fails, instead of throwing an exception

Minor fixes

* Fix segfault when running out of VRAM

Co-authored-by: slaren <slarengh@gmail.com>

---------

Co-authored-by: slaren <slarengh@gmail.com>
 2024-06-11 21:20:29 +02:00
Deven Mistry 14f83526cd
fix broken link in pr template (#7880) [no ci] 
* fix broken link in pr template

* Update pull_request_template.md [no ci]

---------

Co-authored-by: Brian <mofosyne@gmail.com>
 2024-06-12 02:18:58 +10:00
Brian 6fe42d073f
github: move PR template to .github/ root (#7868) 2024-06-11 17:43:41 +03:00
Johannes Gäßler 148995e5e5
llama-bench: more compact markdown tables (#7879) 2024-06-11 14:45:40 +02:00
Georgi Gerganov 4bfe50f741
tests : check the Python version (#7872) 
ggml-ci
 2024-06-11 10:10:20 +03:00
Johannes Gäßler bdcb8f4222
CUDA: int8 tensor cores for MMQ (q4_K, q5_K, q6_K) (#7860) 2024-06-11 08:26:07 +02:00
slaren c2ce6c47e4
fix CUDA CI by using a windows-2019 image (#7861) 
* try to fix CUDA ci with --allow-unsupported-compiler

* trigger when build.yml changes

* another test

* try exllama/bdashore3 method

* install vs build tools before cuda toolkit

* try win-2019
 2024-06-11 08:59:20 +03:00
Olivier Chafik b61eb9644d
json: refine constraint for whitespace to avoid runaways yet allow pretty print (#7866) 2024-06-11 02:22:57 +01:00
Olivier Chafik 396b18dfec
`json`: document schema conversion in GBNF readme, align manual grammar examples & converters (#7841) 
* json: fix char pattern in grammar converters

* json: prevent number precision & whitespace runaways in example grammars

* json: add doc to grammar readme
 2024-06-11 01:00:30 +01:00
Jared Van Bortel 864a99e7a0
cmake : fix CMake requirement for CUDA (#7821) 2024-06-10 18:32:10 -04:00
slaren fd5ea0f897
ci : try win-2019 on server windows test (#7854) 2024-06-10 15:18:41 +03:00
Georgi Gerganov c28a83902c
examples : remove --instruct remnants (#7846) 2024-06-10 15:00:15 +03:00
Georgi Gerganov d9da0e4986
server : improve "prompt" handling (#7847) 2024-06-10 14:59:55 +03:00
Johannes Gäßler 1f0dabda8d
CUDA: use tensor cores for MMQ (#7676) 
* CUDA: int8 tensor cores for MMQ (legacy quants)

* fix out-of-bounds writes

* __builtin_assume -> GGML_CUDA_ASSUME

* fix writeback returning too early
 2024-06-10 11:45:13 +02:00
Ben Ashbaugh af4ae502dd
use the correct SYCL context for host USM allocations (#7777) 
Signed-off-by: Ben Ashbaugh <ben.ashbaugh@intel.com>
 2024-06-10 10:21:31 +01:00
Georgi Gerganov 10ceba354a
flake.lock: Update (#7838) 
Flake lock file updates:

• Updated input 'nixpkgs':
    'github:NixOS/nixpkgs/ad57eef4ef0659193044870c731987a6df5cf56b?narHash=sha256-SzDKxseEcHR5KzPXLwsemyTR/kaM9whxeiJohbL04rs%3D' (2024-05-29)
  → 'github:NixOS/nixpkgs/051f920625ab5aabe37c920346e3e69d7d34400e?narHash=sha256-4q0s6m0GUcN7q%2BY2DqD27iLvbcd1G50T2lv08kKxkSI%3D' (2024-06-07)

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
 2024-06-09 16:04:50 -07:00
Georgi Gerganov e95beeb1fc
imatrix : handle partial entries (#7833) 2024-06-09 20:19:35 +03:00
Nicolás Pérez 57bf62ce7c
docs: Added initial PR template with directions for doc only changes and squash merges [no ci] (#7700) 
This commit adds pull_request_template.md and CONTRIBUTING.md . It focuses on explaining to contributors the need to rate PR complexity level, when to add [no ci] and how to format PR title and descriptions.

Co-authored-by: Brian <mofosyne@gmail.com>
Co-authored-by: compilade <git@compilade.net>
 2024-06-10 01:24:29 +10:00
mgroeber9110 3e2ee44315
server: do not remove whitespace at the start of a completion chunk (#7830) 2024-06-09 20:50:35 +10:00
Johannes Gäßler 42b53d192f
CUDA: revise q8_1 data layout for mul_mat_q (#7824) 2024-06-09 09:42:25 +02:00
sasha0552 2decf57bc6
convert-hf : set the model name based on cli arg, if present (#7693) 
`--model-name` argument was added a while ago but did not do anything.
This commit fixes this issue and enables this feature.
 2024-06-09 16:39:25 +10:00
compilade 5795b94182
convert-hf : match model part name prefix and suffix (#7687) 
In #7075, to fix the conversion of (some) models using model-00001-of-00001.safetensors instead of model.safetensors for a single model part we simply used the same logic as the part count to get the part names. 

But this doesn't always work correctly, like when unusual additional model files like consolidated.safetensors in https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3 are present.

This commit matching both the prefix and the suffix of the model part names should fix this problem without breaking any previously-supported upstream models. But according to report by @teleprint-me there is still some
persistent problem, but shall do in the meantime.
 2024-06-09 12:47:25 +10:00
compilade ed9f252118
gguf-py : decouple adding metadata from writing in GGUFWriter (#7827) 
Main changes of this PR is to consolidate GGUFWriter.add_key and GGUFWriter.add_val into GGUFWriter.add_key_value. 

In addition use_temp_file is now opt-in instead of opt-out defaulting to False.

Also GGUFWriter now does not require output file name until when actually writing to it.

And GGUFWriter doesn't really need to eagerly prepare the data layout of the metadata
 2024-06-09 12:34:29 +10:00
slaren fe1e3917cf
Revert "[SYCL] Update rpc-server.cpp to include SYCL backend (#7682)" (#7808) 
This reverts commit 9422c5e34b.
 2024-06-09 01:43:39 +02:00
Olivier Chafik d4d915d351
url: save -mu downloads to new cache location (#7826) 
* url: save -mu download to new cache location

* url: fs_get_cache_file_path util

* url: tweak sig of fs_get_cache_file
 2024-06-08 21:21:08 +02:00
sasha0552 7a16ce7db2
server : smart slot selection using Longest Common Prefix (#7728) 
* server : Smart selection of available slot using Longest Common Substring

* add usage

* remove trailing whitespaces

* Use Longest Common Prefix (LCP) instead of LCS

* Rename argument
 2024-06-08 10:50:31 +03:00
slaren da799b4189
vulkan : reuse parent extra for views (#7806) 
* vulkan : reuse parent extra for views

* Fix validation error when multiple compute contexts are used in a graph

---------

Co-authored-by: 0cc4m <picard12@live.de>
 2024-06-07 19:47:49 +02:00
Christian Zhou-Zheng c00fad71e5
gguf-split : change binary multi-byte units to decimal (#7803) 2024-06-07 15:56:01 +03:00
intelmatt 27615f5ab2
cmake : fix BUILD_SHARED_LIBS=ON build (#7784) 
common depends on pthreads in Linux
 2024-06-07 15:15:07 +03:00
Johannes Gäßler 7027b27d76
server: update cache_prompt documentation [no ci] (#7745) 2024-06-07 11:15:49 +02:00
woodx a5cabd7649
server : do not get prompt in infill mode (#7286) 
* avoid to get prompt in infill mode and embedding mode

* remove embedding mode

* refactor format

---------

Co-authored-by: wudexiang <wudexiang@bytedance.com>
 2024-06-07 10:09:45 +03:00
pengxin99 d5c938cd77
[SYCL] fix softmax r2r result wrong issue (#7811) 2024-06-07 14:28:26 +08:00
slaren c9ee7118d5
check for nans in imatrix and quantize (#7807) 
* imatrix : detect nan/inf values

* quantize : check imatrix for nan/inf values
 2024-06-07 09:01:29 +03:00