ggerganov/llama.cpp

Pull Requests Commits

Vaibhavs10 committed 59 days ago

99619529

Vaibhavs10 committed 62 days ago

97c64a09

Update the graph.

Vaibhavs10 committed 77 days ago

6201b438

fix errors in conversion.

Vaibhavs10 committed 79 days ago

02ff0850

Model -> ModelBase.

Vaibhavs10 committed 79 days ago

32ea9c5f

Init - first pass.

Vaibhavs10 committed 79 days ago

024bd294

common : suggest --jinja when autodetection fails (#14222)

CISC committed 80 days ago

Verified e434e691

server : fix incorrect usage of llama_get_embeddings() (#14225)

ggerganov committed 80 days ago

Verified 89fea80d

llama : add thread safety test (#14035)

slaren committed 80 days ago

Verified 6adc3c3e

cmake: clean up external project logic for vulkan-shaders-gen (#14179)

bandoti committed 80 days ago

Verified 0dbcabde

model : add NeoBERT (#14164)

huydt84 committed 80 days ago

Verified ad590be9

HIP: disable rocwmma on gfx12 by default until rocm 7.0 (#14202)

IMbackK committed 80 days ago

Verified 7d6d91ba

llama : rework embeddings logic (#14208)

ggerganov committed 80 days ago

Verified d3e64b9f

ggml: Add Android support for GGML_CPU_ALL_VARIANTS (#14206)

chaxu01 committed 80 days ago

Verified 3ba0d843

convert : remove arcee change in convert_hf_to_gguf_update.py (#14207)

bartowski1182 committed 81 days ago

Verified 0bf49eb6

gguf-py : allow key override when adding value to GGUFWriter (#14194)

huydt84 committed 81 days ago

Verified 4ad24367

vulkan: mutex around vkQueueSubmit (#14127)

jeffbolznv committed 81 days ago

Verified c89c2d1a

ggml-cpu : rework weak alias on apple targets (#14146)

xctan committed 81 days ago

Verified 3555b300

model : Add support for Arcee AI's upcoming AFM model (#14185)

bartowski1182 committed 81 days ago

Verified d7da8dc8

server : When listening on a unix domain socket don't print http:// and port (#14180)

ericcurtin committed 81 days ago

Verified cd355eda

quantize : change int to unsigned int for KV overrides (#14197)

EAddario committed 81 days ago

Verified 30e5b01d

CUDA/HIP: fix ssm_scan on devices where warp size is not 32 (#14196)

IMbackK committed 81 days ago

Verified e54b3940

HIP: Replace usage of depricated preprocessor macro __AMDGCN_WAVEFRONT_SIZE__ (#14183)

IMbackK committed 81 days ago

Verified 2c2caa44

kv-cache : fix use-after-move of defrag info (#14189)

ggerganov committed 82 days ago

Verified 5fce5f94

model : add dots.llm1 architecture support (#14044) (#14118)

Noeda committed 82 days ago

Verified 9ae4143b

cparams : rename LLAMA_MAX_PARALLEL_SEQUENCES to LLAMA_MAX_SEQ (#14188)

ggerganov committed 82 days ago

Verified c311ac66

batch : auto-gen positions + verify multi-sequence input (#14177)

ggerganov committed 82 days ago

Verified b9912ac5

docs : remove WIP since PR has been merged (#13912)

pepijndevos committed 82 days ago

Verified 00ba7726

llama-chat : Do not throw when tool parsing fails (#14012)

p1-0tr committed 82 days ago

Verified 3cb203c8

compare-llama-bench: add option to plot (#14169)

am17an committed 82 days ago

Verified 2e42be42

Older