Tags

Tags give the ability to mark specific points in history as being important

b1509

bb50a792 · Add ReLU and SQR CUDA ops to (partially) fix Persimmon offloading (#4041) · Nov 13, 2023
b1505

d96ca7de · server : fix crash when prompt exceeds context size (#3996) · Nov 10, 2023
b1503

4a4fd3ee · server : allow continue edit on completion mode (#3950) · Nov 10, 2023
b1502

df9d1293 · Unbreak persimmon after #3837 (#4010) · Nov 10, 2023
b1500

57ad015d · server : add min_p param (#3877) · Nov 08, 2023
b1499

875fb428 · ggml-alloc : fix backend assignments of views (#3982) · Nov 08, 2023
b1497

413503d4 · make : do not add linker flags when compiling static llava lib (#3977) · Nov 07, 2023
b1496

e9c1cecb · ggml : fix backward rope after YaRN (#3974) · Nov 07, 2023
b1495

54b4df88 · Use params when loading models in llava-cli (#3976) · Nov 07, 2023
b1494

46876d2a · cuda : supports running on CPU for GGML_USE_CUBLAS=ON build (#3946) · Nov 07, 2023
b1493

381efbf4 · llava : expose as a shared library for downstream projects (#3613) · Nov 07, 2023
b1492

2833a6f6 · ggml-cuda : fix f16 mul mat (#3961) · Nov 05, 2023
b1491

d9ccce2e · Allow common process_escapes to handle \x sequences (#3928) · Nov 05, 2023
b1489

132d25b8 · cuda : fix disabling device with --tensor-split 1,0 (#3951) · Nov 05, 2023
b1488

3d48f42e · llama : mark LLM_ARCH_STARCODER as full offload supported (#3945) · Nov 05, 2023
b1487

c41ea36e · cmake : MSVC instruction detection (fixed up #809) (#3923) · Nov 05, 2023
b1486

a7fac013 · ci : use intel sde when ci cpu doesn't support avx512 (#3949) · Nov 05, 2023
b1485

48ade945 · cuda : revert CUDA pool stuff (#3944) · Nov 05, 2023
b1483

d9b33fe9 · metal : round up to 16 to fix MTLDebugComputeCommandEncoder assertion (#3938) · Nov 03, 2023
b1481

abb77e73 · ggml-cuda : move row numbers to x grid dim in mmv kernels (#3921) · Nov 03, 2023

Prev
1
…
46
47
48
49
50
51
52
53
54
…
99
Next