ngxson/llama.cpp

Pull Requests Commits

code aestheticization

christianazinn committed 2 years ago

db3ba108

add command-line args for num threads, num completions file lines, always reload model

christianazinn committed 2 years ago

62560367

fix square_diff matmul index range and CRLF->LF line endings

christianazinn committed 2 years ago

4d7d71bc

fix zero output & param parsing, functional templating

christianazinn committed 2 years ago

4d88cd1a

preliminary template/multiprompt support

christianazinn committed 2 years ago

fa85ba6a

fix matrix transpose multiplication

christianazinn committed 2 years ago

31f153fe

ngxson committed 2 years ago

d446c6d8

ngxson committed 2 years ago

287da25f

add multi prompts, multi-thread for PCA

ngxson committed 2 years ago

447023fc

example template completions

christianazinn committed 2 years ago

dc46264f

param parsing, refactor, comments

christianazinn committed 2 years ago

f58f6af1

proof-of-concept stdlib implementation

christianazinn committed 2 years ago

73747fe8

ngxson committed 2 years ago

b30bea32

ngxson committed 2 years ago

c31c118d

add control-vector-generator

ngxson committed 2 years ago

0a46d730

readme : remove trailing space (#7469)

ggerganov committed 2 years ago

Verified 74f33adf

ggml : silence UB sanitizer error during iq2_xxs quantization (#0)

ggerganov committed 2 years ago

Verified 1debe727

Fix phi3 chat template confusion with zephyr (#7449)

tristandruyen committed 2 years ago

Verified 007489e8

readme : add Bunny in supported models [no ci] (#7469)

criminact committed 2 years ago

Verified 8b94e799

llama : add getters for n_threads/n_threads_batch (#7464)

danbev committed 2 years ago

Verified 3015851c

ci : use Pythia models instead of OpenLlama (#7470)

ggerganov committed 2 years ago

Verified 55ac3b7a

readme : add GPT-NeoX + Pythia to the list of supported models (#7491)

felladrin committed 2 years ago

Verified dacfcebd

Add missing inference support for GPTNeoXForCausalLM (Pythia and GPT-NeoX base models) (#7461)

fairydreaming committed 2 years ago

Verified 9b82476e

llama : rename n_ctx -> cache.size, less confusing (#0)

ggerganov committed 2 years ago

Verified a61a94e5

labeler.yml: add embedding label detector [no ci] (#7482)

mofosyne committed 2 years ago

Verified 152da28a

ggml : remove ggml_flash_attn and ggml_flash_ff (#7463)

ggerganov committed 2 years ago

Verified d48c88cb

ggml : drop support for QK_K=64 (#7473)

ggerganov committed 2 years ago

Verified e84b71c2

Update vulkan rope implementation to support frequency factors (#7475)

0cc4m committed 2 years ago

Verified 1b1e27cb

main : minor (#7462)

ggerganov committed 2 years ago

Verified fbf777d2

CUDA: fix FA out-of-bounds reads (#7479)

JohannesGaessler committed 2 years ago

Verified cd93a28c

Older