ggml-org/llama.cpp

Pull Requests Commits
llama : allow for user specified embedding pooling type (#5849)
iamlemec committed 1 year ago
Verified 475df1d6
gguf-dump : support i-quants (#5841)
Nindaleth committed 1 year ago
Verified 87c2e8b2
llama : fix llama_copy_state_data with fragmented KV cache (#5840)
compilade committed 1 year ago
Verified de9692a7
ci : schedule slow server tests only on Release or on demand (#5839)
phymbert committed 1 year ago
Verified e6029348
server : init http requests thread pool with --parallel if set (#5836)
phymbert committed 1 year ago
Verified 8ef969af
flake.lock: Update (#5842)
ggerganov committed 1 year ago
Verified fa974646
server: tests: passkey challenge / self-extend with context shift demo (#5832)
phymbert committed 1 year ago
Verified 97311342
llama : add abort_callback to interrupt computation (#5409)
Xarbirus committed 1 year ago
Verified 4a6e2d61
ggml : fix IQ3_S AVX implementation (#5834)
ggerganov committed 1 year ago
Verified 494c8703
convert : automatically fall back to HfVocab if tokenizer.model doesn't exist (#5821)
cebtenzzre committed 1 year ago
Verified 4d4d2366
convert-hf : make model class definitions self-contained (#5825)
cebtenzzre committed 1 year ago
Verified c7a0ad8e
ggml : IQ3_S improvements (#5829)
ikawrakow committed 1 year ago
Verified bbde6eb2
scripts : add pod-llama.sh
ggerganov committed 1 year ago
Verified ef2cd694
llama : refactor internal quantization functions (#5830)
ngxson committed 1 year ago
Verified 6c32d8c7
llama : fix segfault from unknown model arch name (#5820)
compilade committed 1 year ago
Verified 802da009
Support multiple GPUs (split mode) on SYCL backend (#5806)
NeoZhangJianyu committed 1 year ago
Verified 71564139
workflows : remove nocleanup arg for check-requirements.sh (#5826)
crasm committed 1 year ago
Verified 9bf297a0
build(nix): Introduce flake.formatter for `nix fmt` (#5687)
ditsuke committed 1 year ago
Verified cb5e8f7f
convert-hf-to-gguf : require einops for InternLM2ForCausalLM (#5792)
Nold360 committed 1 year ago
Verified da3b9ba2
llama : add StarCoder2 support (#5795)
pacman100 committed 1 year ago
Verified c29af7e2
server : remove api_like_OAI.py proxy script (#5808)
ggerganov committed 1 year ago
Verified 38d16b14
ggml-vulkan: fix VULKAN_CHECK_RESULTS flag, which was previously broken (#5813)
ddpasa committed 1 year ago
Verified c2224f00
gemma : fix bfloat16 -> float16 conversion issue (#5810)
kunal-vaishnavi committed 1 year ago
Verified e7433867
common : fix flag `--logits-all` to `--all-logits` (#5805)
ensan-hcl committed 1 year ago
Verified f49a5356
llama : cleanup unused mmq flags (#5772)
phymbert committed 1 year ago
Verified 3ab8b3a9
unicode : switch to multimap based nfd_map (#5799)
iamlemec committed 1 year ago
Verified 9600d59e
server: allow to override threads server pool with --threads-http (#5794)
phymbert committed 1 year ago
Verified 5cb02b4a
ci : add Ubuntu 22 Vulkan CI run (#5789)
netrunnereve committed 1 year ago
Verified 6ea0f010
server : fix newlines in help (#5785)
ggerganov committed 1 year ago
Verified f105471e
[SYCL] Use batched mul_mat pathway (#5591)
AidanBeltonS committed 1 year ago
Verified 38d15216
Terms Of Use Privacy Policy Refund Policy Impressum