ggml-org/llama.cpp

Pull Requests Commits
main : honor --verbose-prompt on interactive prompts (#14350)
CISC committed 177 days ago
Verified abf24104
jinja : Add Mistral-Small-3.2-24B-Instruct-2506.jinja (#14349)
bartowski1182 committed 177 days ago
Verified 901e20bb
CUDA/HIP: optimize mmv paths taken for HIP devices (#14324)
IMbackK committed 177 days ago
Verified 0142961a
ci: add workflow for relocatable cmake package (#14346)
bandoti committed 177 days ago
Verified ce82bd01
vulkan: update windows SDK in release.yml (#14344)
jeffbolznv committed 178 days ago
Verified bf2a99e3
llama : better rwkv chat template and add missing `inputs.use_jinja` setting (#14336)
MollySophia committed 178 days ago
Verified 72c6bc3f
CUDA: mul_mat_v support for batch sizes > 1 (#14262)
JohannesGaessler committed 178 days ago
Verified defe2158
kv-cells : fix tracking of seq_pos (#14339)
ggerganov committed 178 days ago
Verified 7b50d589
vulkan: update windows SDK in CI (#14334)
jeffbolznv committed 178 days ago
Verified 3a9457df
quantize : handle user-defined pruning of whole layers (blocks) (#13037)
EAddario committed 178 days ago
Verified fa4a9f2a
gguf-py : fix SpecialVocab parsing when post_processor is null (#14330)
CISC committed 178 days ago
Verified 238005c2
run : avoid double tokenization (#14327)
retr0reg committed 178 days ago
Verified 66aba7ac
examples : fix is_first logic for tokenization (#14329)
ggerganov committed 178 days ago
Verified f1f5e82d
HIP: enable vec fattn on RDNA4 (#14323)
IMbackK committed 179 days ago
Verified af3373f1
mtmd : fix Pixtral OOM with large images by capping image_size to 1024 (#14326)
yuiseki committed 179 days ago
Verified 5d5c066d
common : use std::string_view now that we target c++17 (#14319)
CISC committed 179 days ago
Verified 40bfa04c
CUDA: add mean operation (#14313)
am17an committed 179 days ago
Verified aa064b2e
gguf-py : fix Qwen3-Embedding eos token (#14314)
CISC committed 179 days ago
Verified aa0ef5c5
Add support for VK_EXT_debug_utils to add labels to Vulkan objects. (#13792)
mtavenrath committed 180 days ago
Verified bb16041c
gguf-py : fix TemplateProcessing pair when bos/eos is missing (#14312)
CISC committed 180 days ago
Verified 58cba76a
metal : fix thread-safety (#14300)
ggerganov committed 180 days ago
Verified 67ae5312
memory : rename interface to llama_memory_context_i (#14296)
ggerganov committed 180 days ago
Verified 692e3cdd
convert : fix Llama 4 conversion (#14311)
danielhanchen committed 180 days ago
Verified b23fa0b3
sync : ggml
ggerganov committed 180 days ago
06cbedfc
Add `ggml_roll` (ggml/1274)
Acly committed 180 days ago
b7147673
docs : fix the link to llama.h (#14293)
david20571015 committed 180 days ago
Verified d860dd99
CUDA: add conv_2d_transpose (#14287)
am17an committed 181 days ago
Verified c959f462
lint : remove trailing whitepace (#14304)
CISC committed 181 days ago
Verified 22015b20
vocab : prevent tokenizer overflow (#14301)
retr0reg committed 181 days ago
Verified dd6e6d0b
sycl: add usage of enqueue_functions extension (#14244)
s-Nick committed 181 days ago
Verified 8308f98c
Terms Of Use Privacy Policy Refund Policy Impressum