ggml-org/llama.cpp

Pull Requests Commits

sycl add conv support

airMeng committed 2 years ago

9aeb0e1f

[SYCL] fix multi-gpu issue on sycl (#8554)

ClarkChin08 committed 2 years ago

Verified ed67bcb2

ggml : add and use ggml_cpu_has_llamafile() (#8664)

ggerganov committed 2 years ago

Verified eddcb523

examples : remove `finetune` and `train-text-from-scratch` (#8669)

ngxson committed 2 years ago

Verified be6d7c07

docs : Quantum -> Quantized (#8666)

Ujjawal-K-Panchal committed 2 years ago

Verified 4b0eff3d

llama: use sliding window for phi3 (#8627)

FanShupei committed 2 years ago

Verified 8a4bad50

readme : update games list (#8673)

MorganRO8 committed 2 years ago

Verified 68504f09

Build Llama SYCL Intel with static libs (#8668)

joeatodd committed 2 years ago

Verified f19bf99c

readme : update UI list [no ci] (#8505)

SommerEngineering committed 2 years ago

Verified 3a7ac530

llama : fix `llama_chat_format_single` for mistral (#8657)

ngxson committed 2 years ago

Verified 96952e71

Re-add erroneously removed -fsycl from GGML_EXTRA_LIBS (#8667)

joeatodd committed 2 years ago

Verified 79167d9e

add llama_lora_adapter_clear (#8653)

ngxson committed 2 years ago

Verified b115105f

examples : Fix `llama-export-lora` example (#8607)

ngxson committed 2 years ago

Verified de280085

server : fix URL.parse in the UI (#8646)

0x4139 committed 2 years ago

Verified b841d074

sycl : Add support for non-release DPC++ & oneMKL (#8644)

joeatodd committed 2 years ago

Verified 64cf50a0

llama : move vocab, grammar and sampling into separate files (#8508)

ggerganov committed 2 years ago

Verified 938943cd

Vulkan IQ4_NL Support (#8613)

0cc4m committed 2 years ago

Verified 751fcfc6

Allow all RDNA2 archs to use sdot4 intrinsic (#8629)

jeroen-mostert committed 2 years ago

Verified 46e47417

contrib : clarify PR squashing + module names (#8630)

ggerganov committed 2 years ago

Verified e7e6487b

[SYCL] fix scratch size of softmax (#8642)

luoyu-intel committed 2 years ago

Verified 063d99ad

llama : fix codeshell support (#8599)

hankeke303 committed 2 years ago

Verified 081fe431

llama : add support for SmolLm pre-tokenizer (#8609)

Stillerman committed 2 years ago

Verified d94c6e0c

*.py: Stylistic adjustments for python (#8233)

jpodivin committed 2 years ago

Verified 566daa5a

llama : allow overrides for tokenizer flags (#8614)

ggerganov committed 2 years ago

Verified 6f11a83e

tests : re-enable tokenizer tests (#8611)

ggerganov committed 2 years ago

Verified e093dd23

llama : add Mistral Nemo inference support (#8604)

iamlemec committed 2 years ago

Verified 50e05353

server : update doc to clarify n_keep when there is bos token (#8619)

kaetemi committed 2 years ago

Verified 62815449

ggml: fix compile error for RISC-V (#8623)

zqb-all committed 2 years ago

Verified 04bab6b7

examples: fix android example cannot be generated continuously (#8621)

devojony committed 2 years ago

Verified b7c11d36

flake.lock: Update (#8610)

ggerganov committed 2 years ago

Verified 45f2c19c

Older