jaime-m-p
|
917dc8cfa6
Tokenizer SPM fixes for phi-3 and llama-spm (#7375)
|
1 жил өмнө |
Georgi Gerganov
|
fabf30b4c4
llama : remove Persimmon (#7408)
|
1 жил өмнө |
Johannes Gäßler
|
20385cebcc
perplexity: update README FP16 results [no ci] (#7413)
|
1 жил өмнө |
Radoslav Gerganov
|
db10f01310
rpc : track allocated buffers (#7411)
|
1 жил өмнө |
Georgi Gerganov
|
3bc10cb485
server : fix temperature + disable some tests (#7409)
|
1 жил өмнө |
AidanBeltonS
|
6bf9b66fa3
[SYCL] Update SYCL upscale operation (#7321)
|
1 жил өмнө |
Bingan
|
26cd4237bc
Update README.md (#7410)
|
1 жил өмнө |
Herman Semenov
|
213e90ed73
ggml-opencl, llama: using reserve() if count already known (#7272)
|
1 жил өмнө |
junchao-loongson
|
65c58207ec
ggml : add loongarch lsx and lasx support (#6454)
|
1 жил өмнө |
Georgi Gerganov
|
1cc0155d04
server : tuning tests (#7388)
|
1 жил өмнө |
Georgi Gerganov
|
e932094d58
server : return error on too large embedding input (#7389)
|
1 жил өмнө |
Georgi Gerganov
|
2789baf480
tests : fix --keep_split -> --keep-split (#7374)
|
1 жил өмнө |
Srihari-mcw
|
33c8d50acc
Add provisions for windows support for BF16 code including CMake provision for enabling AVX512_BF16 (#7258)
|
1 жил өмнө |
slaren
|
d359f30921
llama : remove MPI backend (#7395)
|
1 жил өмнө |
Fred Douglas
|
1ea2a0036e
quantize : fix --keep-split check (#7374)
|
1 жил өмнө |
0cc4m
|
f030ec1f7a
Vulkan Embedding Fix (#7360)
|
1 жил өмнө |
slaren
|
e4e6f67be6
ggml : fix another case of quants nans (#7387)
|
1 жил өмнө |
Johannes Gäßler
|
5ca49cbecd
ggml: implement quantized KV cache for FA (#7372)
|
1 жил өмнө |
Johannes Gäßler
|
1b01f06db0
server: add test for token probs (#7347)
|
1 жил өмнө |
Johannes Gäßler
|
41858392e1
server: fix seed being reported back (#7382)
|
1 жил өмнө |
Anas Ahouzi
|
6aade19ee7
Add StableLM2 pre-tokenizer (#7349)
|
1 жил өмнө |
slaren
|
ab33f7a338
cuda : clear error after buffer allocation failure (#7376)
|
1 жил өмнө |
Brian
|
e23b974f4c
labeler.yml: Use settings from ggerganov/llama.cpp [no ci] (#7363)
|
1 жил өмнө |
Georgi Gerganov
|
854d365aba
cmake : update android comments (#7341)
|
1 жил өмнө |
fraxy-v
|
f5bf761747
Capture CUDA logging output (#7298)
|
1 жил өмнө |
Georgi Gerganov
|
059031b8c4
ci : re-enable sanitizer runs (#7358)
|
1 жил өмнө |
Georgi Gerganov
|
511182eabb
android : use "ci-android" branch for CI (#7341)
|
1 жил өмнө |
Johannes Gäßler
|
133d99c599
CUDA: deduplicate FlashAttention code (#7352)
|
1 жил өмнө |
Johannes Gäßler
|
cb42c29427
server: correct --threads documentation [no ci] (#7362)
|
1 жил өмнө |
Engininja2
|
d233b507cd
cuda : add half2 __shfl_xor() for ROCm 5.5 (#7263)
|
1 жил өмнө |