 蕭澧邦
|
cddae4884c
Correct typo run_llama2.sh > run-llama2.sh (#9149)
|
il y a 1 an |
 tc-mb
|
7ea8d80d53
llava : the function "clip" should be int (#9237)
|
il y a 1 an |
 Faisal Zaghloul
|
42c76d1358
Threadpool: take 2 (#8672)
|
il y a 1 an |
 Jan Boon
|
9f7d4bcf5c
server : fix crash when error handler dumps invalid utf-8 json (#9195)
|
il y a 1 an |
 Georgi Gerganov
|
1d1ccce676
flake.lock: Update (#9162)
|
il y a 1 an |
 slaren
|
9fe94ccac9
docker : build images only once (#9225)
|
il y a 1 an |
|
slaren
|
66b039a501
docker : update CUDA images (#9213)
|
il y a 1 an |
|
Georgi Gerganov
|
20f1789dfb
vulkan : fix build (#0)
|
il y a 1 an |
|
Georgi Gerganov
|
231cff5f6f
sync : ggml
|
il y a 1 an |
 Xie Yanbo
|
3246fe84d7
Fix minicpm example directory (#9111)
|
il y a 1 an |
 compilade
|
78eb487bb0
llama : fix qs.n_attention_wv for DeepSeek-V2 (#9156)
|
il y a 1 an |
 Xuan Son Nguyen
|
a77feb5d71
server : add some missing env variables (#9116)
|
il y a 1 an |
 CausalLM
|
2e59d61c1b
llama : fix ChatGLM4 wrong shape (#9194)
|
il y a 1 an |
 Carsten Kragelund Jørgensen
|
75e1dbbaab
llama : fix llama3.1 rope_freqs not respecting custom head_dim (#9141)
|
il y a 1 an |
 arch-btw
|
ad76569f8e
common : Update stb_image.h to latest version (#9161)
|
il y a 1 an |
|
slaren
|
7d787ed96c
ggml : do not crash when quantizing q4_x_x with an imatrix (#9192)
|
il y a 1 an |
|
Georgi Gerganov
|
06658ad7c3
metal : separate scale and mask from QKT in FA kernel (#9189)
|
il y a 1 an |
|
Georgi Gerganov
|
fc18425b6a
ggml : add SSM Metal kernels (#8546)
|
il y a 1 an |
|
Georgi Gerganov
|
879275ac98
tests : fix compile warnings for unreachable code (#9185)
|
il y a 1 an |
|
Georgi Gerganov
|
7a3df798fc
ci : add VULKAN support to ggml-ci (#9055)
|
il y a 1 an |
|
Georgi Gerganov
|
e5edb210cd
server : update deps (#9183)
|
il y a 1 an |
|
slaren
|
0c41e03ceb
metal : gemma2 flash attention support (#9159)
|
il y a 1 an |
|
slaren
|
f12ceaca0c
ggml-ci : try to improve build time (#9160)
|
il y a 1 an |
 Justine Tunney
|
436787f170
llama : fix time complexity of string replacement (#9163)
|
il y a 1 an |
 Herman Semenov
|
93bc3839f9
common: fixed not working find argument --n-gpu-layers-draft (#9175)
|
il y a 1 an |
 Johannes Gäßler
|
f91fc5639b
CUDA: fix Gemma 2 numerical issues for FA (#9166)
|
il y a 1 an |
|
Johannes Gäßler
|
e11bd856d5
CPU/CUDA: Gemma 2 FlashAttention support (#8542)
|
il y a 1 an |
 João Dinis Ferreira
|
8f824ffe8e
quantize : fix typo in usage help of `quantize.cpp` (#9145)
|
il y a 1 an |
|
Xuan Son Nguyen
|
3ba780e2a8
lora : fix llama conversion script with ROPE_FREQS (#9117)
|
il y a 1 an |
 piDack
|
a07c32ea54
llama : use F32 precision in GLM4 attention and no FA (#9130)
|
il y a 1 an |
