Skip to content
GitLab
Explore
Sign in
Tags
Tags give the ability to mark specific points in history as being important
b2194
13e2c771
·
cmake : remove obsolete sycl compile flags (#5581)
·
Feb 19, 2024
b2193
f53119ce
·
minor : fix trailing whitespace (#5538)
·
Feb 19, 2024
b2191
4480542b
·
baby-llama : allocate graphs in ggml_context (#5573)
·
Feb 19, 2024
b2190
11b12de3
·
llama : add llama_chat_apply_template() (#5538)
·
Feb 19, 2024
b2189
3a9cb4ca
·
cuda, metal : fix nans in soft_max (#5574)
·
Feb 19, 2024
b2187
f0d1fafc
·
ggml : android and old glibc NUMA incompatibility bugfixes (#5557)
·
Feb 19, 2024
b2186
a0c2dad9
·
build : pass all warning flags to nvcc via -Xcompiler (#5570)
·
Feb 18, 2024
b2185
14278f55
·
ggml : restore vec dot stride arg names (#5453)
·
Feb 18, 2024
b2184
b1de9682
·
ci : fix wikitext url + compile warnings (#5569)
·
Feb 18, 2024
b2182
5ee99c32
·
common, server : surface min_keep as its own parameter (#5567)
·
Feb 18, 2024
b2181
c145f8a1
·
server : slots monitoring endpoint (#5550)
·
Feb 18, 2024
b2180
689a091b
·
sampling : do not set min_keep to n_probs (#5564)
·
Feb 18, 2024
b2179
f3f28c53
·
cmake : fix GGML_USE_SYCL typo (#5555)
·
Feb 18, 2024
b2178
e75c6279
·
server : enhanced health endpoint (#5548)
·
Feb 18, 2024
b2177
36376abe
·
server : --n-predict option document and cap to max value (#5549)
·
Feb 18, 2024
b2176
66c1968f
·
server : graceful server shutdown (#5244)
·
Feb 18, 2024
b2175
1dcc3fde
·
common : fix ub (#5530)
·
Feb 18, 2024
b2174
5d3de51f
·
ggml, common, examples, tests : fixed type arguments in printf (#5528)
·
Feb 18, 2024
b2172
bd2d4e39
·
1.5 bit quantization (#5453)
·
Feb 18, 2024
b2167
5bf2b94d
·
cmake : fix VULKAN and ROCm builds (#5525)
·
Feb 16, 2024
Prev
1
…
24
25
26
27
28
29
30
31
32
…
99
Next