Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update to use llama.cpp/master-aacdbd4 #8

Open
wants to merge 355 commits into
base: v2
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
355 commits
Select commit Hold shift + click to select a range
ea3a0ad
llama : update stubs for systems without mmap and mlock (#1266)
xloem May 1, 2023
58b367c
cuBLAS: refactor and optimize f16 mat mul performance (#1259)
slaren May 1, 2023
f4cef87
Add git-based build information for better issue tracking (#1232)
DannyDaemonic May 1, 2023
2d099e5
ggml: add names to tensors (#1268)
slaren May 2, 2023
e2cd506
main : switch input_noecho to input_echo to remove negation (#979)
deadprogram May 2, 2023
2bb992f
llama : allow 0 as a seed number. (#1275)
rbrisita May 2, 2023
cc0bb72
ggml : fix ppc64le build error and make cmake detect Power processors…
mgiessing May 2, 2023
8c9be35
examples : improve vertical alignment of a few variables (#1286)
deadprogram May 2, 2023
5d5817c
ggml : fix 32-bit ARM
ggerganov May 2, 2023
0e6cbff
llama : fix compile warnings
ggerganov May 2, 2023
67c7779
examples : add llama_init_from_gpt_params() common function (#1290)
deadprogram May 2, 2023
bf4b22f
fix missing parameters in `llama_init_from_gpt_params` (#1293)
slaren May 2, 2023
9daff41
fix build-info.h for git submodules (#1289)
kuvaus May 3, 2023
55bc5f0
Call sh on build-info.sh (#1294)
DannyDaemonic May 3, 2023
13b0c68
Handle signals properly on Windows (#1123)
DannyDaemonic May 3, 2023
2485d7a
Process escape sequences given in prompts (#1173)
DannyDaemonic May 3, 2023
e216aa0
llama : only copy used KV cache in get / set state (#1272)
ejones May 3, 2023
a8a2efd
examples : various prompt and example fixes (#1298)
CRD716 May 3, 2023
b0c71c7
scripts : platform independent script to verify sha256 checksums (#1203)
KASR May 3, 2023
e2a937c
minor : fix trailing whitespaces
ggerganov May 3, 2023
bca9ad9
minor : fix whitespaces (#1302)
ggerganov May 3, 2023
6daa09d
examples : read chat prompts from a template file (#1196)
khimaros May 3, 2023
799fdc1
ggml : vectorize Q8_0 quantization
ggerganov May 3, 2023
f647ce0
fix #1224 reverse prompt and multi line (#1297)
newTomas May 4, 2023
c65a7fb
Update main's README.md with new features (#1296)
DannyDaemonic May 4, 2023
db10808
Only escape prompts when used with `-e` (#1311)
DannyDaemonic May 4, 2023
20fbf2a
ggml : change immintrin.h to intrin.h for compatibility (#1307)
iRonJ May 4, 2023
2edbdb0
main : add --in-suffix option (#1318)
44670 May 4, 2023
360cfe5
readme : add OpenBuddy link (#1321)
44670 May 4, 2023
d3e8093
convert: support DT_BF16 tensors (#1309)
ivanstepanovftw May 4, 2023
34d9f22
Wrap exceptions in std::exception to verbose output on exception. (#1…
ivanstepanovftw May 4, 2023
94c5652
quantize: make output filename optional, default to ggml-model-<ftype…
slaren May 4, 2023
a90e96b
Convert.py @staticmethod (#1327)
blecaillon May 5, 2023
2d13786
Fix for OpenCL / clbast builds on macOS. (#1329)
IonoclastBrigham May 5, 2023
921dcee
readme: add missing info (#1324)
prusnak May 5, 2023
a3b85b2
ci : add cublas to windows release (#1271)
Green-Sky May 5, 2023
173d0e6
makefile: automatic Arch Linux detection (#1332)
daniandtheweb May 5, 2023
3924088
Remove default arguments from sampling functions (#1343)
j-f1 May 6, 2023
1b0fd45
ggml : Allow usage of CLBlast alongside Accelerate.framework (#1336)
swittk May 7, 2023
e129551
CI: add Windows CLBlast and OpenBLAS builds (#1277)
SlyEcho May 7, 2023
1f48b0a
Documented CUDA reproducibility, added warning (#1346)
JohannesGaessler May 8, 2023
95078cc
convert: add ability to convert safetensors files (#1276)
ubik2 May 8, 2023
f9a6364
llama : require first token to be BOS (#1303)
ggerganov May 8, 2023
003ba2f
llama : fix hparams shadow (#1367)
prusnak May 8, 2023
fe60904
readme : add TOC and Pygmalion instructions (#1359)
AlpinDale May 8, 2023
56551bc
readme : add notice about upcoming breaking change
ggerganov May 8, 2023
41654ef
Interface improvements and `--multiline-input` (previously `--author-…
DannyDaemonic May 9, 2023
9f8dbc4
use pause asm insn in busyloop to run the CPU (13600K) 10 °C cooler (…
Safari77 May 9, 2023
e6a46b0
Locale fix for Windows (#1379)
DannyDaemonic May 9, 2023
cf348a6
main : add option to save full output to session (#1338)
ejones May 10, 2023
b608b55
prompts : model agnostic DAN (#1304)
CRD716 May 11, 2023
b9fd7ee
ggml : remove bit shuffling (#1405)
ggerganov May 11, 2023
089b1c9
readme : add C#/.NET bindings repo (#1409)
AsakusaRinne May 12, 2023
553fd4d
Add clang-tidy reviews to CI (#1407)
slaren May 12, 2023
773ee24
CLI args use - instead of _, backwards compatible (#1416)
JohannesGaessler May 12, 2023
fb62f92
llama : fix --mtest option (close #1414)
ggerganov May 12, 2023
699b1ad
opencl : fix kernels for the new formats (#1422)
SlyEcho May 13, 2023
738ace3
llama : free ggml context in set / copy state data (close #1425)
ggerganov May 13, 2023
cdd5350
readme : update Q4_0 perplexities
ggerganov May 13, 2023
6456a4e
embedding : remove unused code (#1426)
AsakusaRinne May 13, 2023
0cd22e1
llama : fix various warnings
ggerganov May 13, 2023
ac0cd25
Adding SSE instructions to ggml_vec_dot_q4_0_q8_0 (#1413)
3ooabkhxtn May 13, 2023
f048af0
ggml : sync alibi fix from ggml repo
ggerganov May 13, 2023
f954edd
ggml : implement backward pass for llama + small training-llama-from-…
xaedes May 13, 2023
905d87b
ggml : GPU-accelerated token generation (#1412)
JohannesGaessler May 13, 2023
66841fd
ggml : multi-thread mul and diag_mask ops (#1428)
ggerganov May 13, 2023
5a5aeb1
llama : fix unused warning
ggerganov May 13, 2023
bda4d7c
make : fix PERF build with cuBLAS
ggerganov May 13, 2023
08737ef
cuda : fix convert function (#1412)
ggerganov May 13, 2023
601a033
ggml : add GGML_QNT_VERSION to track quantization format changes
ggerganov May 14, 2023
60f8c36
ggml : add AVX support based on AVX2 code (#1430)
katsu560 May 14, 2023
13c351a
ggml : various fixes (#1450)
ggerganov May 14, 2023
79b2d5b
ggml : alternative fix for race condition bug in non-inplace ggml_com…
xaedes May 14, 2023
eb36362
cuda : deduplicated dequantization code (#1453)
JohannesGaessler May 14, 2023
b5c9295
benchmark-matmul: fix clang-tidy issues, report results in GFLOPS (#1…
slaren May 14, 2023
63d2046
fix get_num_physical_cores() (#1436)
zrm May 15, 2023
2a5ee02
Add alternate include path for openblas (#1476)
sandyiscool May 16, 2023
9560655
define default model path once, sync path with readme (#1366)
ott2 May 16, 2023
4262742
~7% faster Q5_1 AVX2 code (#1477)
ilyakurdyukov May 16, 2023
2b26469
convert.py: Support models which are stored in a single pytorch_model…
TheBloke May 16, 2023
c238b58
benchmark-matmul: Print the average of the test results (#1490)
rankaiyx May 17, 2023
dc271c5
Remove unused n_parts parameter (#1509)
sw May 17, 2023
ee96541
Fixes #1511 lambda issue for w64devkit (mingw) (#1513)
DannyDaemonic May 18, 2023
5ea4339
make kv_f16 the default for api users (#1517)
Green-Sky May 18, 2023
4b7e245
minor : fix compile warnings
ggerganov May 19, 2023
79e3efb
readme : adds WizardLM to the list of supported models (#1485)
dakennedyd May 19, 2023
7694b52
main : make reverse prompt option act as a stop token in non-interact…
data-angel May 19, 2023
943e608
examples : add persistent chat (#1495)
ejones May 19, 2023
6986c78
tests : add missing header
ggerganov May 19, 2023
2d5db48
ggml : use F16 instead of F32 in Q4_0, Q4_1, Q8_0 (#1508)
ggerganov May 19, 2023
4fd3e29
ggml : fix scalar implementation of Q4_1 dot
ggerganov May 20, 2023
8a203f9
llama : fix compile warnings in llama_set_state_data()
ggerganov May 20, 2023
503db28
llama : fix name shadowing and C4146 (#1526)
maximegmd May 20, 2023
d2c59b8
Fix for mingw (#1462)
DannyDaemonic May 20, 2023
ec2e10c
llama : add llama_init_backend() API (close #1527)
ggerganov May 20, 2023
07e9ace
feature : add blis and other BLAS implementation support (#1502)
zenixls2 May 20, 2023
ea60007
Revert "feature : add blis and other BLAS implementation support (#15…
ggerganov May 20, 2023
affc76e
cuda : loading models directly into VRAM, norm calculation on GPU, br…
JohannesGaessler May 20, 2023
3de84b2
ggml : add ggml_clamp() (#1539)
ggerganov May 20, 2023
29cf559
llama : define magic numbers as integer constants (#1518) (#1520)
imaami May 20, 2023
9ecb30f
OpenCL: Fixes for older devices. (#1435)
SlyEcho May 20, 2023
b8ee340
feature : support blis and other blas implementation (#1536)
zenixls2 May 20, 2023
fab49c6
ggml : update WASM SIMD
ggerganov May 20, 2023
265db98
ggml : output 3d sizes in ggml_graph_dump_dot()
ggerganov May 21, 2023
7780e4f
make : .PHONY clean (#1553)
stsydow May 21, 2023
7e4ea5b
examples : add server example with REST API (#1443)
FSSRepo May 21, 2023
2e6cd4b
OpenCL Token Generation Acceleration (#1459)
0cc4m May 22, 2023
7d87381
Fix handling of "invalid property" when creating OpenCL command queue…
mthuurne May 23, 2023
1359b6a
chat-persistent.sh : use bracket expressions in grep (#1564)
Senemu May 24, 2023
c31bbe9
readme : add docs for chat-persistent.sh (#1568)
ejones May 24, 2023
ac7876a
Update CLBlast to 1.6.0 (#1580)
SlyEcho May 24, 2023
1fcdcc2
cuda : performance optimizations (#1530)
JohannesGaessler May 25, 2023
66874d4
Some improvements to loading the session with --prompt-cache (#1550)
KerfuffleV2 May 26, 2023
bdbda1b
ggml : sync ggml core (minor additions, e.g. ggml_get_tensor_by_name())
ggerganov May 27, 2023
83c54e6
[CI] CLBlast: Fix directory name (#1606)
SlyEcho May 27, 2023
9361803
ggml : add ggml_tensor_overhead()
ggerganov May 27, 2023
0ecb1bb
[CI] Fix openblas (#1613)
SlyEcho May 27, 2023
97c9b77
Add documentation about CLBlast (#1604)
SlyEcho May 27, 2023
0df7d63
Include server in releases + other build system cleanups (#1610)
KerfuffleV2 May 27, 2023
a670464
ggml : add support for the RISCV architecture (#1616)
apcameron May 27, 2023
ca74884
opencl : use strstr to check if fp16 supported (#1611)
howard0su May 28, 2023
bb051d9
opencl : no need to allocate cl_mem on heap (#1612)
howard0su May 28, 2023
337aea1
examples : add --alias option to gpt_params to set use friendly model…
epicfilemcnulty May 28, 2023
1b78ed2
Only show -ngl option when relevant + other doc/arg handling updates …
KerfuffleV2 May 28, 2023
3b126f6
LLAMA_DEBUG adds debug symbols (#1617)
JohannesGaessler May 28, 2023
0e730dd
Adding git in container package dependencies (#1621)
jpodivin May 29, 2023
2483676
Work around for recalculating logits in cached prompts (Fixes #1585) …
DannyDaemonic May 29, 2023
5d1830b
ggml : fix bug in ggml_alibi
ggerganov May 29, 2023
7552ac5
ggml : sync cgraph import / export API
ggerganov May 29, 2023
ffb06a3
OpenLLaMA 3B support (#1588)
SlyEcho May 30, 2023
136476e
Fix prompt cache saving and chat-persistent rollover (#1678)
ejones Jun 3, 2023
b5c8546
Docker: change to calling convert.py (#1641)
jpodivin Jun 3, 2023
d8bd001
Add info about CUDA_VISIBLE_DEVICES (#1682)
SlyEcho Jun 3, 2023
dcb2ed4
OpenCL: Fix duplication of layers in VRAM and RAM, add GPU mul kernel…
0cc4m Jun 4, 2023
ecb217d
llama : Metal inference (#1642)
ggerganov Jun 4, 2023
827f5ed
readme : update hot topics
ggerganov Jun 4, 2023
d1f563a
llama : fix Metal KV cache sync (close #1695)
ggerganov Jun 5, 2023
5220a99
Increase 3B scratch buffers. (#1698)
SlyEcho Jun 5, 2023
99009e7
ggml : add SOTA 2,3,4,5,6 bit k-quantizations (#1684)
ikawrakow Jun 5, 2023
e7fe66e
ci : disable auto tidy (#1705)
ggerganov Jun 5, 2023
efe0507
ggml : fix internal overflow in ggml_time_us on Windows (#1702)
grahameth Jun 5, 2023
9d0693b
metal : use shared buffers between CPU and GPU (#1696)
kiltyj Jun 5, 2023
c2df36d
llama : consistently catch and throw only exceptions deriving from st…
mgroeber9110 Jun 5, 2023
f146562
readme : fix typo (#1700)
Foul-Tarnished Jun 5, 2023
f4c55d3
docs : add performance troubleshoot + example benchmark documentation…
Yuval-Peled Jun 5, 2023
590250f
metal : add checks for buffer size (#1706)
spencersutton Jun 6, 2023
7a74dee
llama : temporary disable Q6_K output quantization (#1711)
ggerganov Jun 6, 2023
7ad7750
gitignore : add .clang-tidy
ggerganov Jun 6, 2023
2d43387
ggml : fix builds, add ggml-quants-k.o (close #1712, close #1710)
ggerganov Jun 6, 2023
d5b111f
Clblast fixes + enhancements to save VRAM and offload more layers (#1…
LostRuins Jun 6, 2023
44f906e
metal : add f16 support
ggerganov Jun 6, 2023
17366df
Multi GPU support, CUDA refactor, CUDA scratch buffer (#1703)
JohannesGaessler Jun 6, 2023
2a4e41a
llama : fix compile warnings
ggerganov Jun 6, 2023
2d7bf11
llama : fix vram_scratch var
ggerganov Jun 6, 2023
35a8491
main: add the possibility to open the prompt cache read-only (#1640)
wtarreau Jun 7, 2023
4dc62c5
readme : add June roadmap
ggerganov Jun 7, 2023
5b57a5b
flake : update to support metal on m1/m2 (#1724)
jpetrucciani Jun 7, 2023
5c64a09
k-quants : allow to optionally disable at compile time (#1734)
ggerganov Jun 7, 2023
0035858
k-quants : add missing compile definition to CMakeLists (#1748)
johnson442 Jun 8, 2023
4161bdc
metal : add Q4_K implementation (#1733)
ikawrakow Jun 8, 2023
53aba3f
clang-tidy : restore dot file from accidental deletion
ggerganov Jun 8, 2023
b50b570
ggml : fix fprintf warnings (#1720)
sroussey Jun 8, 2023
8fc8179
Add llama.cpp docker support for non-latin languages (#1673)
qingfengfenga Jun 8, 2023
0f291e1
metal : Q6_K implementation (#1752)
ikawrakow Jun 8, 2023
8432d4d
ggml : load data into int8x16x4_t using vld4q_s8 on arm64 (#1738)
lindeer Jun 8, 2023
0bf7cf1
Revert "ggml : load data into int8x16x4_t using vld4q_s8 on arm64 (#1…
ggerganov Jun 8, 2023
72ff528
metal : add Q2_K implementation (#1762)
ikawrakow Jun 8, 2023
245fc3c
metal : faster q4_0 (#1775)
ikawrakow Jun 9, 2023
92f44ff
metal : add GELU implementation (#1770)
manyoso Jun 9, 2023
b33dee2
metal : fix build "tanhf" -> "tanh"
ggerganov Jun 9, 2023
ae9663f
Windows nvcc workaround (#1753)
JohannesGaessler Jun 9, 2023
98ed165
OpenCL: Add release memory (#1741)
edp1096 Jun 9, 2023
555275a
make : add SSSE3 compilation use case (#1659)
rankaiyx Jun 10, 2023
ef3171d
ggml : workaround for missing _mm256_setr_m128i in GCC < 8 (#1638)
xingchensong Jun 10, 2023
4f0154b
llama : support requantizing models instead of only allowing quantiza…
KerfuffleV2 Jun 10, 2023
e9b66ee
metal : add Q4_1 implementation (#1785)
ikawrakow Jun 10, 2023
17c10ac
ggml : force no_alloc == false when creating opt tensors (close #1699)
ggerganov Jun 10, 2023
059e990
doc : fix wrong address of BLIS.md (#1772)
Aisuko Jun 10, 2023
303f580
metal : fix issue with ggml-metal.metal path. Closes #1769 (#1782)
abetlen Jun 10, 2023
3f12231
k-quants : GCC12 compilation fix (#1792)
vagran Jun 10, 2023
4de0334
cmake : fix Metal build (close #1791)
ggerganov Jun 10, 2023
31d2b5f
Update SHA256SUMS with current hashes for models quantized using q4_0…
rlanday Jun 11, 2023
12b063f
Fixed WSL cuda's OOM error (#1594)
JoelSeniorLiang Jun 11, 2023
fa84c4b
Fix issue where interactive mode crashes when input exceeds ctx size …
KerfuffleV2 Jun 11, 2023
8c0a10e
metal : fix failure to load model (#1817)
ikawrakow Jun 12, 2023
58970a4
Leverage mmap for offloading tensors to GPU (#1597)
howard0su Jun 12, 2023
e4caa8d
ci : run when changing only the CUDA sources (#1800)
slaren Jun 12, 2023
74a6d92
Metal implementation for all k_quants (#1807)
ikawrakow Jun 12, 2023
74d4cfa
Allow "quantizing" to f16 and f32 (#1787)
KerfuffleV2 Jun 13, 2023
2347e45
llama : do a warm-up eval at start for better timings (#1824)
ggerganov Jun 13, 2023
e32089b
train : improved training-from-scratch example (#1652)
xaedes Jun 13, 2023
9254920
baby-llama : fix operator!= (#1821)
0xspringtime Jun 13, 2023
254a7a7
CUDA full GPU acceleration, KV cache in VRAM (#1827)
JohannesGaessler Jun 14, 2023
6b8312e
Better error when using both LoRA + GPU layers (#1861)
JohannesGaessler Jun 15, 2023
4bfcc85
metal : parallel command buffer encoding (#1860)
ggerganov Jun 15, 2023
64cc19b
Fix the validation of main device (#1872)
howard0su Jun 15, 2023
37e257c
make : clean *.so files (#1857)
sandyiscool Jun 15, 2023
9dda13e
readme : server compile flag (#1874)
nivibilla Jun 15, 2023
cf267d1
make : add train-text-from-scratch (#1850)
daboe01 Jun 15, 2023
69b34a0
swift : Package compile breaks due to ggml-metal.metal (#1831)
Schaltfehler Jun 15, 2023
3559433
cmake : set include path for OpenBlas (#1830)
okigan Jun 15, 2023
c36e81d
examples : add chat-vicuna.sh (#1854)
yangli2 Jun 15, 2023
bed9275
cmake : remove whitespaces
ggerganov Jun 15, 2023
a09f919
Fixed CUDA runtime version check (#1879)
JohannesGaessler Jun 15, 2023
602c748
gitignore : add several entries specific to Visual Studio (#1888)
iboB Jun 16, 2023
3d01122
CUDA : faster k-quant dot kernels (#1862)
ikawrakow Jun 16, 2023
9cbf50c
build : fix and ignore MSVC warnings (#1889)
iboB Jun 16, 2023
5b9ccaf
Fixed possible macro redefinition (#1892)
FrankHB Jun 16, 2023
ac3b886
llama : fix embd when offloading non-repeating layers (#1891)
JohannesGaessler Jun 16, 2023
13fe9d2
cmake : add auto detection of BLAS_INCLUDE_DIRS (#1886)
zenixls2 Jun 16, 2023
b41b4ca
examples : add "simple" (#1840)
SuperUserNameMan Jun 16, 2023
d411968
opencl : support k-quants (#1836)
0cc4m Jun 16, 2023
92f20d9
train : get raw text instead of page with html (#1905)
Davidy22 Jun 17, 2023
b4c6f46
Allow cmake to build ggml as a library (#1896)
KerfuffleV2 Jun 17, 2023
bac1992
readme : alternative way to build for Android with CLBlast. (#1828)
gustrd Jun 17, 2023
5ddf7ea
hooks : setting up flake8 and pre-commit hooks (#1681)
jpodivin Jun 17, 2023
794db3e
Server Example Refactor and Improvements (#1570)
digiwombat Jun 17, 2023
fc45a81
exposed modules so that they can be invoked by nix run github:ggergan…
faezs Jun 17, 2023
0711a5f
metal : add norm, cpy f16->f16, alibi kernels (#1823)
apage43 Jun 17, 2023
3d59ec5
ggml : fix warnings under MSVC (#1908)
howard0su Jun 17, 2023
86c7571
make : update for latest Arch (#1701)
daniandtheweb Jun 17, 2023
051e1b0
llama : fix kv_cache `n` init (close #1903)
ggerganov Jun 17, 2023
2c9380d
Only one CUDA stream per device for async compute (#1898)
JohannesGaessler Jun 17, 2023
4f9c43e
minor : warning fixes
ggerganov Jun 17, 2023
b241649
make : do not print help for simple example
ggerganov Jun 17, 2023
57cd694
cmake : add CUDA_ARCHITECTURES to new target ggml_static (#1917)
howard0su Jun 18, 2023
ce2c7d7
metal : handle buffers larger than device's maxBufferLength (#1826)
ggerganov Jun 18, 2023
90cc59d
examples : fix examples/metal (#1920)
ikawrakow Jun 18, 2023
8ab8ba6
llama : prevent usage of k-quants when tensor size is not a multiple …
ikawrakow Jun 18, 2023
e1886cf
readme : update Android build instructions (#1922)
mikeyang01 Jun 18, 2023
8596af4
ggml : fix bug in ggml_compute_forward_add_q_f32 (#1918)
l3utterfly Jun 18, 2023
0ede372
Fixed incorrectly applying RMS norm twice (#1925)
JohannesGaessler Jun 18, 2023
b24c304
Added tokens per second to info prints (#1928)
JohannesGaessler Jun 18, 2023
16b9cd1
Convert vector to f16 for dequantize mul mat vec (#1913)
JohannesGaessler Jun 19, 2023
1e3abfc
cmake : fix build shared ggml when CUDA is enabled (#1929)
howard0su Jun 19, 2023
b97ca43
ggml : sync latest ggml repo (#1924)
ggerganov Jun 19, 2023
ca7c3f4
cuda : faster k-quants on older GPUs (#1930)
ikawrakow Jun 19, 2023
cb40dfc
llama : only use Q6_K for output weights if tensor size is multiple o…
ikawrakow Jun 19, 2023
23fc5c2
cmake : fix trailing whitespaces
ggerganov Jun 19, 2023
ba4e85a
llama : use aligned memory during ggml_init call from loading saved s…
l3utterfly Jun 19, 2023
18b3562
ggml : fix bug in LBFGS optimizer (found by ggml tests)
ggerganov Jun 19, 2023
20568fe
[Fix] Reenable server embedding endpoint (#1937)
SlyEcho Jun 19, 2023
aacdbd4
llama : fix params struct slignment (#1936)
mudler Jun 20, 2023
de1cd08
Merge tag 'master-aacdbd4' into v2
alexrozanski Jun 20, 2023
d12a031
remove scripts/ dir
alexrozanski Jun 20, 2023
d153e95
move metal sources to llamaObjCxx sources dir
alexrozanski Jun 20, 2023
0c7a9f1
remove ggml-opencl
alexrozanski Jun 20, 2023
3021d8d
fix build after merge from master-aacdbd4
alexrozanski Jun 20, 2023
75090a9
add missing llama_tokenize()
alexrozanski Jun 20, 2023
9f55b61
fix llama_tokenize() implementation
alexrozanski Jun 20, 2023
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
20 changes: 20 additions & 0 deletions .github/workflows/tidy-post.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,20 @@
name: clang-tidy review post comments

on:
workflow_dispatch:
workflows: ["clang-tidy-review"]
types:
- completed

jobs:
build:
runs-on: ubuntu-latest

steps:
- uses: ZedThree/clang-tidy-review/[email protected]
# lgtm_comment_body, max_comments, and annotations need to be set on the posting workflow in a split setup
with:
# adjust options as necessary
lgtm_comment_body: ''
annotations: false
max_comments: 25
23 changes: 23 additions & 0 deletions .github/workflows/tidy-review.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,23 @@
name: clang-tidy-review

on:
pull_request:
branches:
- master

jobs:
clang-tidy-review:
runs-on: ubuntu-latest

steps:
- uses: actions/checkout@v3

- uses: ZedThree/[email protected]
id: review
with:
lgtm_comment_body: ''
build_dir: build
cmake_command: cmake . -B build -DCMAKE_EXPORT_COMPILE_COMMANDS=on
split_workflow: true

- uses: ZedThree/clang-tidy-review/[email protected]
22 changes: 21 additions & 1 deletion .gitignore
Original file line number Diff line number Diff line change
@@ -1,14 +1,22 @@
*.o
*.a
.DS_Store
.build/
.cache/
.direnv/
.envrc
.swiftpm
.venv
.clang-tidy
.vs/
.vscode/
.DS_Store

models/*
*.bin

arm_neon.h
compile_commands.json
CMakeSettings.json

# Xcode
#
Expand Down Expand Up @@ -100,3 +108,15 @@ fastlane/test_output
# https://github.com/johnno1962/injectionforxcode

iOSInjectionProject/
=======
__pycache__

zig-out/
zig-cache/

ppl-*.txt
qnt-*.txt
perf-*.txt

examples/jeopardy/results.txt
>>>>>>> master-aacdbd4
1 change: 1 addition & 0 deletions Package.swift
Original file line number Diff line number Diff line change
Expand Up @@ -28,6 +28,7 @@ let package = Package(
name: "llamaObjCxx",
dependencies: [],
path: "Sources/llamaObjCxx",
exclude: ["cpp/ggml-metal.metal"],
publicHeadersPath: "headers",
cSettings: [.unsafeFlags(["-Wno-shorten-64-to-32"]), .define("GGML_USE_ACCELERATE")],
cxxSettings: [
Expand Down
Loading