ngxson/llama.cpp

Pull Requests Commits

Apply suggestions from code review

ngxson committed 2 years ago

Verified f7320036

add GGML_ASSERT

ngxson committed 2 years ago

11b1564e

ngxson committed 2 years ago

540c3016

move batch_allocr inside decode/encode_internal

ngxson committed 2 years ago

6ab116ac

llama : fix empty batch cause llama_batch_allocr to crash

ngxson committed 2 years ago

bd697ca7

readme : update bindings list (#9951)

lcarrere committed 2 years ago

Verified 45f09764

readme : update infra list (#9942)

icppWorld committed 2 years ago

Verified 7cab2083

llama : remove all_pos_0, all_pos_1, all_seq_id from llama_batch (#9745)

ngxson committed 2 years ago

Verified cda0e4b6

rpc : backend refactoring (#9912)

rgerganov committed 2 years ago

Verified afd9909a

[SYCL] Add SYCL Backend registry, device and Event Interfaces (#9705)

OuadiElfarouki committed 2 years ago

Verified 87421a23

add amx kernel for gemm (#8998)

mingfeima committed 2 years ago

Verified 60ce97c9

server : add n_indent parameter for line indentation requirement (#9929)

ggerganov committed 2 years ago

Verified 8901755b

llama : rename batch_all to batch (#8881)

danbev committed 2 years ago

Verified 6f55bccb

readme : remove --memory-f32 references (#9925)

ggerganov committed 2 years ago

Verified 17bb9280

llama : change warning to debug log

ggerganov committed 2 years ago

Verified 9f45fc1e

llama : infill sampling handle very long tokens (#9924)

ggerganov committed 2 years ago

Verified 99bd4ac2

readme : update bindings list (#9918)

ShenghaiWang committed 2 years ago

Verified 3752217e

vulkan : add backend registry / device interfaces (#9721)

slaren committed 2 years ago

Verified f010b77a

fix: allocating CPU buffer with size `0` (#9917)

giladgd committed 2 years ago

Verified 21942002

fix: use `vm_allocate` to allocate CPU backend buffer on macOS (#9875)

giladgd committed 2 years ago

Verified 73afe681

llama : suppress conversion from 'size_t' to 'int' (#9046)

danbev committed 2 years ago

Verified 9e041024

llava : fix typo in error message [no ci] (#9884)

danbev committed 2 years ago

Verified dbf18e4d

grammar : fix JSON Schema for string regex with top-level alt. (#9903)

jemc committed 2 years ago

Verified 66c2c930

llama : add tensor name for "result_norm" (#9907)

MollySophia committed 2 years ago

Verified 10433e8b

server : fix the disappearance of the end of the text (#9867)

z80maniac committed 2 years ago

Verified 1f66b699

ggerganov committed 2 years ago

Verified 0e41b300

ggml-alloc : remove buffer_id from leaf_alloc (ggml/987)

danbev committed 2 years ago

Verified cd60b88b

[CANN] Fix cann compilation error (#9891)

leo-pony committed 2 years ago

Verified becfd387

llama : add infill sampler (#9896)

ggerganov committed 2 years ago

Verified 755a9b2b

server : improve infill context reuse (#9894)

ggerganov committed 2 years ago

Verified 223c25a7

Older