ggerganov/llama.cpp

Pull Requests Commits

server : add `n_discard` parameter (#6300)

kaetemi committed 1 year ago

3d032ece

nix: make `xcrun` visible in Nix sandbox for precompiling Metal shaders (#6118)

josephst committed 1 year ago

e190f1fc

cuda : rename build flag to LLAMA_CUDA (#6299)

slaren committed 1 year ago

28034596

nix: fix blas support (#6281)

ck3d committed 1 year ago

b06c16ef

tests : include IQ2_XXS and IQ2_XS in test-quantize-fns (#6303)

ikawrakow committed 1 year ago

1f2fd4e7

flake.lock: Update (#6266)

ggerganov committed 1 year ago

43139cc5

cuda : fix LLAMA_CUDA_F16 build (#6298)

slaren committed 1 year ago

2f34b865

cuda : refactor into multiple files (#6269)

slaren committed 1 year ago

ae1f211c

Server: clean up OAI params parsing function (#6284)

ngxson committed 1 year ago

ad3a0505

[SYCL] fix SYCL backend build on windows is break by LOG() error (#6290)

NeoZhangJianyu committed 1 year ago

95ad616c

examples : add "retrieval" (#6193)

mscheong01 committed 1 year ago

64e7b47c

ggml : support AVX512VNNI (#6280)

jart committed 1 year ago

7733f0c7

Fix heap corruption from wmode out-of-bound writes on windows (#6272)

TheFlipbook committed 1 year ago

a32b77c4

imatrix : fix wname for mul_mat_id ops (#6271)

ggerganov committed 1 year ago

a0e584de

Fixed lookup compilation issues on Windows (#6273)

JohannesGaessler committed 1 year ago

7aed0ffe

ci : close inactive issue, increase operations per run (#6270)

phymbert committed 1 year ago

ea279d56

sampling : deduplicated code for probability distribution access (#6240)

mscheong01 committed 1 year ago

586e7bc5

[SYCL] offload op (#6217)

airMeng committed 1 year ago

ddf65685

Support build win release for SYCL (#6241)

NeoZhangJianyu committed 1 year ago

d03224ac

use _wfopen instead of fopen on Windows (#6248)

cebtenzzre committed 1 year ago

94d1b3b4

gitignore : gguf-split

ggerganov committed 1 year ago

95562175

common: llama_load_model_from_url split support (#6192)

phymbert committed 1 year ago

f482bb2e

server: docs: `--threads` and `--threads`, `--ubatch-size`, `--log-disable` (#6254)

phymbert committed 1 year ago

1997577d

llama : add grok-1 support (#6204)

arki05 committed 1 year ago

476b0251

split: add gguf-split in the make build target (#6262)

phymbert committed 1 year ago

21cad01b

server: flush stdout after logging in both text and json layout (#6253)

phymbert committed 1 year ago

1b26aebe

lookup: complement data from context with general text statistics (#5479)

JohannesGaessler committed 1 year ago

50ccaf5e

common : default --hf-file to --model (#6234)

ggerganov committed 1 year ago

56a00f0a

convert-llama2c-to-ggml : enable conversion of GQA models (#6237)

fraxy-v committed 1 year ago

92397d87

quantize: options for output and token embedding tensors qtype (#6239)

ikawrakow committed 1 year ago

1d0331c1

Newer Older