ggml-org/llama.cpp

Pull Requests Commits
ggml : increased GGML_MAX_PARAMS to allow finetuning of 70b models (#4424)
Taikono-Himazin committed 1 year ago
Verified 41a11aaf
Update README.md (#4388)
y10ab1 committed 1 year ago
Verified 8a7b2fa5
grammar : revert the replacement of llama_token_to_piece with id_to_token (#4396)
AlienKevin committed 1 year ago
Verified e18f7345
sync : ggml (new ops, tests, backend, etc.) (#4359)
ggerganov committed 1 year ago
Verified fe680e3d
llama : per-layer KV cache + quantum K cache (#4309)
ggerganov committed 1 year ago
Verified bcc0eb45
train : fix #4227 (double free in examples/train-text-from-scratch/train-text-from-scratch.cpp) (#4351)
casavaca committed 1 year ago
Verified 81bc9214
server : recognize cache_prompt parameter in OAI API (#4347)
ggerganov committed 1 year ago
Verified 05cd6e50
common : fix compile warning
ggerganov committed 1 year ago
Verified caa92492
speculative : support `--color` (#4343)
stduhpf committed 1 year ago
Verified da5eaef1
grammar : pre-computed pieces + reserve mem + less string copies (#4330)
MarcusDunn committed 1 year ago
Verified 5f6e0c0d
llama : allow overriding GGUF metadata when loading model (#4092)
KerfuffleV2 committed 1 year ago
Verified 5aa365d8
sampling : custom samplers order (#4285)
MaggotHATE committed 1 year ago
Verified 52c8bc3c
swift : revert compiler checks for swift package (#4332)
kchro3 committed 1 year ago
Verified e4b76bbe
simple : update error message for KV cache check (#4324)
danbev committed 1 year ago
Verified 23b5e12e
swift : fix concatenation method to avoid invalid UTF8 stringfication (#4325)
ensan-hcl committed 1 year ago
Verified d208995c
swift : fix prompt tokenization logic (#4321)
ensan-hcl committed 1 year ago
Verified 5c9f90cb
grammar-parser : fix typo (#4318)
eltociear committed 1 year ago
Verified 4fa44e84
ggml : reuse ggml_get_n_tasks() in ggml_graph_plan() (#4308)
ggerganov committed 1 year ago
Verified fbbc4282
ggml : fix soft max out-of-bounds access (#4307)
ggerganov committed 1 year ago
Verified adf3de4f
server : fix OpenAI API `stop` field to be optional (#4299)
Mardak committed 1 year ago
Verified 33e171d1
py : add grammar to oai like api (#4294)
neph1 committed 1 year ago
Verified 6949b50d
llama : pad KV cache size (#4280)
ggerganov committed 1 year ago
Verified d7b800b8
llama : avoid using "optional" keyword (#4283)
ggerganov committed 1 year ago
Verified 5a7d3125
llama : support optional tensors (#4283)
ggerganov committed 1 year ago
Verified d5a1cbde
swift : fix token_to_piece implementation (#4278)
ensan-hcl committed 1 year ago
Verified b220222a
build : enable libstdc++ assertions for debug builds (#4275)
cebtenzzre committed 1 year ago
Verified 511f52c3
llama : support attention bias on LLaMA architecture (#4283)
RealJosephus committed 1 year ago
Verified 03562f3a
llama : add Qwen support (#4281)
simonJJJ committed 1 year ago
Verified 37c746d6
llama : fix integer overflow during quantization (#4284)
ggerganov committed 1 year ago
Verified 880f5797
py : add requirements file for convert-hf-to-gguf.py (#4277)
danbev committed 1 year ago
Verified 8d6d9f03
Terms Of Use Privacy Policy Refund Policy Impressum