ggerganov/llama.cpp

Pull Requests Commits
llama : fix defrag bugs + add parameter (#5735)
ggerganov committed 2 years ago
Verified 9d533a77
Makefile: use variables for cublas (#5689)
lindeer committed 2 years ago
Verified cbbd1efa
fix server hangs on empty prompt (#5733)
ngxson committed 2 years ago
Verified b11a93df
Adding IQ2_S and IQ2_M to complete coverage of the 2-3 bit quantization range (#5721)
ikawrakow committed 2 years ago
Verified a33e6a0d
CUDA: fix DEBUG_CUDA_MALLOC (#5729)
JohannesGaessler committed 2 years ago
Verified 47bb7b48
readme : update ui list (#5731)
guinmoon committed 2 years ago
Verified c4d7f817
[SYCL] Add support for soft_max ALiBi (#5639)
AidanBeltonS committed 2 years ago
Verified e849078c
unicode : reuse iterator (#5726)
ggerganov committed 2 years ago
Verified 67fd3313
server: CI fix trailing space (#5728)
phymbert committed 2 years ago
Verified 4804215c
server: CI tests reduce build matrix (#5725)
phymbert committed 2 years ago
Verified 8a533f0d
llama : fix Gemma rope type (#5691)
ggerganov committed 2 years ago
Verified 269de86b
flake.lock: Update
github-actions[bot] committed 2 years ago
c3937339
server: tests - slow inference causes timeout on the CI (#5715)
phymbert committed 2 years ago
Verified e3965cf3
server: docs - refresh and tease a little bit more the http server (#5718)
phymbert committed 2 years ago
Verified 8b350356
llama : refactor k-shift implementation + KV defragmentation (#5691)
ggerganov committed 2 years ago
Verified bf08e006
server : fix crash when system prompt is bigger than batch size (#5714)
compilade committed 2 years ago
Verified f7625019
ggml-quants : provide ggml_vqtbl1q_u8 for 64bit compatibility (#5711)
rgryta committed 2 years ago
Verified abbabc5e
make : fix nvcc version is empty (#5713)
kwin1412 committed 2 years ago
Verified f1a98c52
readme : add Msty to UI list (#5618)
ashokgelal committed 2 years ago
Verified 7d548a18
server: logs - unified format and --log-format option (#5700)
phymbert committed 2 years ago
Verified 930b1780
server: concurrency fix + monitoring - add /metrics prometheus compatible endpoint (#5708)
phymbert committed 2 years ago
Verified d52d7819
cmake : fix compilation for Android armeabi-v7a (#5702)
rgryta committed 2 years ago
Verified 12894088
code : normalize enum names (#5697)
ggerganov committed 2 years ago
Verified ab336a9d
py : fix StableLM conversion after config.json changes (#5703)
aahouzi committed 2 years ago
Verified 69917dfa
server: continue to update other slots on embedding concurrent request (#5699)
phymbert committed 2 years ago
Verified 9e359a4f
IQ3_S: a much better alternative to Q3_K (#5676)
ikawrakow committed 2 years ago
Verified 4c4cb307
server: init functional tests (#5566)
phymbert committed 2 years ago
Verified 525213d2
server : add KV cache quantization options (#5684)
AlpinDale committed 2 years ago
Verified fd43d66f
convert : fix missing ftype for gemma (#5690)
cebtenzzre committed 2 years ago
Verified 54fbcd2c
mpt : do not duplicate token_embd.weight on disk (#5670)
cebtenzzre committed 2 years ago
Verified 15499eb9
Terms Of Use Privacy Policy Refund Policy Impressum