ggml-org/llama.cpp

Pull Requests Commits
make rms_norm_eps a parameter (#2374)
slaren committed 2 years ago
Verified 41c67416
Chat UI extras (#2366)
akx committed 2 years ago
Verified b3f138d0
ggml : sync (unary ops refactor, static-correctness) (#2370)
ggerganov committed 2 years ago
Verified 5b2b2dc6
Fix scalar version of Q5_K when QK_K = 64 (#2362)
ikawrakow committed 2 years ago
Verified 42f70cb2
llama : add grammar-based sampling (#1773)
ejones committed 2 years ago
Verified 84e09a7d
Some more Q4_K and Q5_K speedup on CUDA (#2346)
ikawrakow committed 2 years ago
Verified 2f9cf974
Add gqa parameter support to the server (#2351)
IgnacioFDM committed 2 years ago
Verified 4f06592c
Fix __dp4a documentation (#2348)
JohannesGaessler committed 2 years ago
Verified 70d26ac3
common : n_threads == -1 uses std::thread::hardware_concurrency() (#2347)
Freed-Wu committed 2 years ago
Verified 57921ca6
fix n_tasks (#2342)
slaren committed 2 years ago
Verified 3602ac42
ggml: move op parameters from tensors to ggml_tensor::op_params (#2333)
slaren committed 2 years ago
Verified 95a6c595
llama : grouped-query attention + LLaMAv2 70B support (#2276)
ggerganov committed 2 years ago
Verified e76d630d
llama : print help to stdout (#2338)
maddes8cht committed 2 years ago
Verified 1d0824b2
flake : support `nix build '.#opencl'` (#2337)
Freed-Wu committed 2 years ago
Verified bc3ec2cd
llama : print max tensor size to stderr (#2336)
Christian Demsar committed 2 years ago
Verified a940458e
make : fix CLBLAST compile support in FreeBSD (#2331)
yukiteruamano committed 2 years ago
Verified 91171b80
examples : simplify vim plugin (#2327)
AustinMroz committed 2 years ago
Verified 355c80f4
metal : support bcast add & dup & cont op (#2323)
li-plus committed 2 years ago
Verified 83a00ce6
Speed up Q4_K (#2322)
ikawrakow committed 2 years ago
Verified d2a43664
CUDA: Fixed 7b q3_K_S with mul_mat_vec_q (#2313)
JohannesGaessler committed 2 years ago
Verified b9b7d94f
llama : optimize memory buffers (#2325)
ggerganov committed 2 years ago
Verified b47b8a9c
Perplexity: Compute scores correlated to HellaSwag (#2312)
klosax committed 2 years ago
Verified b5fe67f8
examples : basic VIM plugin
whoreson committed 2 years ago
Verified 24baa54a
ci : fix args
ggerganov committed 2 years ago
Verified dd6c67d3
ci : add 7B CUDA tests (#2319)
ggerganov committed 2 years ago
Verified 5d500e8c
examples : add easy python script to create quantized (k-bit support) GGML models from local HF Transformer models (#2311)
richardr1126 committed 2 years ago
Verified 7d5f1846
Custom RoPE + bettter memory management for CUDA (#2295)
ikawrakow committed 2 years ago
Verified d924522a
Faster Q3_K implementation on Metal (#2307)
ikawrakow committed 2 years ago
Verified 4d76a5f4
ggml : fix the rope fix (513f8619535a64fa9ace808cdcbcf66211535f5c)
ggerganov committed 2 years ago
Verified 0db14fef
examples : fix typo in minigpt4.py (#2298)
eltociear committed 2 years ago
Verified 03e56697
Terms Of Use Privacy Policy Refund Policy Impressum