ggerganov/llama.cpp

Pull Requests Commits
ggml : add ggml-common.h to deduplicate shared code (#5940)
ggerganov committed 1 year ago
Verified 8a3012a4
server : simplify logic for empty prompts (#5953)
ggerganov committed 1 year ago
Verified 9674aaf3
Server: reorganize some http logic (#5939)
ngxson committed 1 year ago
Verified 950ba1ab
server : add SSL support (#5926)
gabe-l-hart committed 1 year ago
Verified e1fa9569
server: tests: add truncated prompt tests, better kv cache size (#5933)
phymbert committed 1 year ago
Verified fd72d2d2
llama : support Mamba Selective State Space Models (#5328)
compilade committed 1 year ago
Verified c2101a2e
llama : fix quantization of shared token_embd (#5944)
compilade committed 1 year ago
Verified 515f7d0d
server: metrics: add llamacpp:prompt_seconds_total and llamacpp:tokens_predicted_seconds_total, reset bucket only on /metrics. Fix values cast to int. Add Process-Start-Time-Unix header. (#5937)
phymbert committed 1 year ago
Verified 76e86882
llama : assume tied weights if lm_head/output weights is missing (#5824)
dmahurin committed 1 year ago
Verified e457fb35
server : fix EOS token detection with disabled cache (#5938)
ggerganov committed 1 year ago
Verified af37fd8b
log : fix MSVC compile errors (#5643)
uextm committed 1 year ago
Verified 581ed5c4
llama-bench : add embeddings option (#5924)
ggerganov committed 1 year ago
Verified 6cdabe65
Revert "[SYCL] fix error when set main gpu to non-zero (#5901)" (#5918)
NeoZhangJianyu committed 1 year ago
Verified 89fb735f
server : add `/v1/completions` endpoint (#5914)
mscheong01 committed 1 year ago
Verified 55a2a900
server : refactor (#5882)
ggerganov committed 1 year ago
Verified 2002bc96
[SYCL] fix error when set main gpu to non-zero (#5901)
NeoZhangJianyu committed 1 year ago
Verified ceca1aef
ggml : use SYS_get_cpu if SYS_getcpu is not defined (#5906)
cebtenzzre committed 1 year ago
Verified e04e04f8
ggml : use `uint8x16_t` return type for `ggml_vqtbl1q_u8` (#5894)
bobqianic committed 1 year ago
Verified e25fb4b1
convert : remove AWQ remnants (#5768)
ggerganov committed 1 year ago
Verified 1e35d619
add wait() to make code stable (#5895)
NeoZhangJianyu committed 1 year ago
Verified 8ced9f7e
compare-llama-bench.py : remove mul_mat_q (#5892)
slaren committed 1 year ago
Verified 652ca2bd
quants : use MM256_SET_M128I consistently to fix gcc 7 build (#5889)
cebtenzzre committed 1 year ago
Verified bd836944
grammars : blacklists character control set (#5888)
ExtReMLapin committed 1 year ago
Verified 3de31677
Revert "grammars : don't allow to output unescaped new line in string (#5885)"
ggerganov committed 1 year ago
Verified 82cb31eb
grammars : don't allow to output unescaped new line in string (#5885)
ExtReMLapin committed 1 year ago
Verified b1a4e994
Vulkan Improvements (#5835)
0cc4m committed 1 year ago
Verified 61d1c88e
[SYCL] fix mul_mat fault in CI/unit-test (#5862)
NeoZhangJianyu committed 1 year ago
Verified 21b08674
fix editorconfig check break (#5879)
mscheong01 committed 1 year ago
Verified 6a87ac3a
fix speculative decoding build on windows (#5874)
jquesnelle committed 1 year ago
Verified 29eee404
nix: static build (#5814)
hutli committed 1 year ago
Verified 1d41d6f7
Terms Of Use Privacy Policy Refund Policy Impressum