Guy Goldenberg
|
3cfbbdb44e
Merge commit from fork
|
il y a 7 mois |
Georgi Gerganov
|
80709b70a2
batch : add LLAMA_BATCH_DEBUG environment variable (#14172)
|
il y a 7 mois |
ddpasa
|
26ff3685bf
docs : Update multimodal.md (#14122)
|
il y a 7 mois |
Georgi Gerganov
|
60c666347b
batch : rework llama_batch_allocr (#14153)
|
il y a 7 mois |
Georgi Gerganov
|
b7cc7745e3
readme : remove survey link (#14168)
|
il y a 7 mois |
Christian Kastner
|
cc8d081879
cmake: Add ability to pass in LLAMA_BUILD_NUMBER/COMMIT (#14167)
|
il y a 7 mois |
Đinh Trọng Huy
|
d714dadb57
pooling : make cls_b and cls_out_b optional (#14165)
|
il y a 7 mois |
Georgi Gerganov
|
ffad043973
server : fix SWA condition for full context reprocess (#14163)
|
il y a 7 mois |
Anton Mitkov
|
0889eba570
sycl: Adding additional cpy dbg print output (#14034)
|
il y a 7 mois |
Ewan Crawford
|
c61285e739
SYCL: Bump oneMath commit (#14152)
|
il y a 7 mois |
Christian Kastner
|
09cf2c7c65
cmake : Improve build-info.cpp generation (#14156)
|
il y a 7 mois |
Georgi Gerganov
|
c33fe8b8c4
vocab : prevent heap overflow when vocab is too small (#14145)
|
il y a 7 mois |
Anton Mitkov
|
ed52f3668e
sycl: Remove not needed copy f16->f32 for dnnl mul mat (#14125)
|
il y a 7 mois |
Georgi Gerganov
|
a681b4ba83
readme : remove project status link (#14149)
|
il y a 7 mois |
Georgi Gerganov
|
7d516443dd
server : re-enable SWA speculative decoding (#14131)
|
il y a 7 mois |
Georgi Gerganov
|
f6e1a7aa87
context : simplify output counting logic during decode (#14142)
|
il y a 7 mois |
Georgi Gerganov
|
c3ee46fab4
batch : remove logits_all flag (#14141)
|
il y a 7 mois |
Georgi Gerganov
|
e2c0b6e46a
cmake : handle whitepsaces in path during metal build (#14126)
|
il y a 7 mois |
Georgi Gerganov
|
9596506965
kv-cache : fix split_equal handling in unified implementation (#14130)
|
il y a 7 mois |
compilade
|
a20b2b05bc
context : round n_tokens to next multiple of n_seqs when reserving (#14140)
|
il y a 7 mois |
bandoti
|
2e89f76b7a
common: fix issue with regex_escape routine on windows (#14133)
|
il y a 7 mois |
Christian Kastner
|
532802f938
Implement GGML_CPU_ALL_VARIANTS for ARM (#14080)
|
il y a 7 mois |
Sigbjørn Skjæret
|
d4e0d95cf5
chore : clean up relative source dir paths (#14128)
|
il y a 7 mois |
Sigbjørn Skjæret
|
cc66a7f78f
tests : add test-tokenizers-repo (#14017)
|
il y a 7 mois |
Jeff Bolz
|
bd248d4dc7
vulkan: Better thread-safety for command pools/buffers (#14116)
|
il y a 7 mois |
Aman
|
7781e5fe99
webui: Wrap long numbers instead of infinite horizontal scroll (#14062)
|
il y a 7 mois |
Georgi Gerganov
|
89a184fa71
kv-cache : relax SWA masking condition (#14119)
|
il y a 7 mois |
Taylor
|
2baf07727f
server : pass default --keep argument (#14120)
|
il y a 7 mois |
Georgi Gerganov
|
7ae2932116
kv-cache : add LLAMA_KV_CACHE_DEBUG environment variable (#14121)
|
il y a 7 mois |
Jeff Bolz
|
1f7d50b293
vulkan: Track descriptor pools/sets per-context (#14109)
|
il y a 7 mois |