 Douglas Hanley
|
03bf161eb6
llama : support batched embeddings (#5466)
|
hai 1 ano |
 Johannes Gäßler
|
ad014bba97
make: add error message for bad CUDA version (#5444)
|
hai 1 ano |
 Georgi Gerganov
|
49cc1f7d67
bert : add tests + fix quantization (#5475)
|
hai 1 ano |
|
Georgi Gerganov
|
99b8b43d7b
tests : disable moe test (#5473)
|
hai 1 ano |
 Kawrakow
|
895407f31b
ggml-quants : fix compiler warnings (shadow variable) (#5472)
|
hai 1 ano |
|
Georgi Gerganov
|
099afc6274
llama : fix quantization when tensors are missing (#5423)
|
hai 1 ano |
|
Georgi Gerganov
|
df334a1125
swift : package no longer use ggml dependency (#5465)
|
hai 1 ano |
 Lee
|
dbd8828eb0
py : fix persimmon `n_rot` conversion (#5460)
|
hai 1 ano |
 Abhilash Majumder
|
43fe07c1a4
ggml-sycl: Replace 3d ops with macro (#5458)
|
hai 1 ano |
 Daniel Bevenius
|
4a46d2b792
llava : remove prog parameter from ArgumentParser (#5457)
|
hai 1 ano |
|
Georgi Gerganov
|
3b169441df
sync : ggml (#5452)
|
hai 1 ano |
|
Johannes Gäßler
|
3bdc4cd0f5
CUDA: mul_mat_vec_q tiling, refactor mul mat logic (#5434)
|
hai 1 ano |
|
Douglas Hanley
|
2891c8aa9a
Add support for BERT embedding models (#5423)
|
hai 1 ano |
 github-actions[bot]
|
97a336507e
flake.lock: Update
|
hai 1 ano |
 Sergio López
|
c88c74f967
vulkan: only use M-sized matmul on Apple GPUs (#5412)
|
hai 1 ano |
 Alexey Parfenov
|
a803333a4e
common : use enums for sampler types (#5418)
|
hai 1 ano |
|
Alexey Parfenov
|
684780141a
server : allow to specify tokens as strings in logit_bias (#5003)
|
hai 1 ano |
|
Georgi Gerganov
|
85910c5b30
main : ctrl+C print timing in non-interactive mode (#3873)
|
hai 1 ano |
|
Georgi Gerganov
|
139b62a839
common : fix compile warning
|
hai 1 ano |
|
Georgi Gerganov
|
0f2411f154
ggml : fix compile warnings (unused vars) (#4966)
|
hai 1 ano |
 snadampal
|
a07d0fee1f
ggml : add mmla kernels for quantized GEMM (#4966)
|
hai 1 ano |
|
Johannes Gäßler
|
e4640d8fdf
lookup: add print for drafting performance (#5450)
|
hai 1 ano |
 Xuan Son Nguyen
|
907e08c110
server : add llama2 chat template (#5425)
|
hai 1 ano |
 Ian Bull
|
f026f8120f
metal : use autoreleasepool to avoid memory leaks (#5437)
|
hai 1 ano |
|
Georgi Gerganov
|
cd9aea63b5
scripts : update sync scripts with new backends
|
hai 1 ano |
|
Georgi Gerganov
|
43b65f5eb8
sync : ggml
|
hai 1 ano |
 Michael Podvitskiy
|
4633d93af0
ggml : add abort_callback for cpu backend (ggml/725)
|
hai 1 ano |
 Neuman Vong
|
4b7b38bef5
vulkan: Set limit for task concurrency (#5427)
|
hai 1 ano |
|
Daniel Bevenius
|
e00d2a62dd
llava : add requirements.txt and update README.md (#5428)
|
hai 1 ano |
 Riley Stewart
|
7c777fcd5d
server : fix prompt caching for repeated prompts (#5420)
|
hai 1 ano |
