Skip to content
GitLab
Explore
Sign in
Tags
Tags give the ability to mark specific points in history as being important
b2576
06957477
·
[Model] Add support for xverse (#6301)
·
Mar 29, 2024
b2573
80939870
·
cmake : add explicit metal version options (#6370)
·
Mar 29, 2024
b2568
be55134a
·
convert : refactor vocab selection logic (#6355)
·
Mar 28, 2024
b2567
66ba5602
·
llava : fix MobileVLM (#6364)
·
Mar 28, 2024
b2566
0308f5e3
·
llama : fix command-r inference when omitting outputs (#6367)
·
Mar 28, 2024
b2563
6902cb7f
·
server : stop gracefully on SIGTERM (#6348)
·
Mar 28, 2024
b2554
25f4a613
·
[SYCL] fix set main gpu crash (#6339)
·
Mar 28, 2024
b2548
e82f9e2b
·
[SYCL] Fix batched impl for NVidia GPU (#6164)
·
Mar 27, 2024
b2543
0642b22c
·
server: public: use relative routes for static files (#6325)
·
Mar 27, 2024
b2542
a4f569e8
·
[SYCL] fix no file in win rel (#6314)
·
Mar 27, 2024
b2541
32c8486e
·
wpm : portable unicode tolower (#6305)
·
Mar 26, 2024
b2540
557410b8
·
llama : greatly reduce output buffer memory usage (#6122)
·
Mar 26, 2024
b2536
deb72401
·
embedding : adjust `n_ubatch` value (#6296)
·
Mar 26, 2024
b2534
e190f1fc
·
nix: make `xcrun` visible in Nix sandbox for precompiling Metal shaders (#6118)
·
Mar 25, 2024
b2531
1f2fd4e7
·
tests : include IQ2_XXS and IQ2_XS in test-quantize-fns (#6303)
·
Mar 25, 2024
b2529
2f34b865
·
cuda : fix LLAMA_CUDA_F16 build (#6298)
·
Mar 25, 2024
b2527
ad3a0505
·
Server: clean up OAI params parsing function (#6284)
·
Mar 25, 2024
b2526
95ad616c
·
[SYCL] fix SYCL backend build on windows is break by LOG() error (#6290)
·
Mar 25, 2024
b2523
a32b77c4
·
Fix heap corruption from wmode out-of-bound writes on windows (#6272)
·
Mar 24, 2024
b2521
7aed0ffe
·
Fixed lookup compilation issues on Windows (#6273)
·
Mar 24, 2024
Prev
1
…
11
12
13
14
15
16
17
18
19
…
98
Next