Sākums Izpētīt Palīdzība
Pierakstīties
cturan
/
llama.cpp
spogulis no https://github.com/cturan/llama.cpp
1
0
Atdalīts 0
Faili Problēmas 0 Vikivietne
Koks: 36d19a603b
Atzari Tagi

Revīziju vēsture

Autors SHA1 Ziņojums Datums
  Stephan Walter 36d19a603b Remove Q4_3 which is no better than Q5 (#1218) 2 gadi atpakaļ
  Georgi Gerganov 55390bcaf2 ggml : sync ggml (ggml_alibi) 2 gadi atpakaļ
  0cc4m 7296c961d9 ggml : add CLBlast support (#1164) 2 gadi atpakaļ
  Georgi Gerganov 574406dc7e ggml : add Q5_0 and Q5_1 quantization (#1187) 2 gadi atpakaļ
  Georgi Gerganov 7a32fcb3b2 ggml : add Q8_0 quantization format (rename the old one to Q8_1) (ARM NEON) (#1179) 2 gadi atpakaļ
  Georgi Gerganov 8a0f8673ba ggml : export symbols (#1155) 2 gadi atpakaļ
  Georgi Gerganov 12b5900dbc ggml : sync ggml (add GPT-NeoX RoPE implementation) 2 gadi atpakaļ
  Kawrakow 38de86a711 llama : multi-threaded quantization (#1075) 2 gadi atpakaļ
  Georgi Gerganov e0305ead3a ggml : add Q4_3 quantization (#1082) 2 gadi atpakaļ
  slaren 8944a13296 Add NVIDIA cuBLAS support (#1044) 2 gadi atpakaļ
  Georgi Gerganov 77a73403ca ggml : add new Q4_2 quantization (ARM only) (#1046) 2 gadi atpakaļ
  slaren 315a95a4d3 Add LoRA support (#820) 2 gadi atpakaļ
  Ivan Komarov f266259ad9 Speedup the AVX-512 implementation of ggml_vec_dot_q4_0() (#933) 2 gadi atpakaļ
  Georgi Gerganov e95b6554b4 ggml : add Q8_0 quantization for intermediate results (#951) 2 gadi atpakaļ
  Pavol Rusnak c56b715269 Expose type name from ggml (#970) 2 gadi atpakaļ
  Kerfuffle c9a59b70a5 ggml : add unary and binary map operations (#874) 2 gadi atpakaļ
  Georgi Gerganov a3a2a0eda8 ggml : add GGML_DEFAULT_N_THREADS 2 gadi atpakaļ
  Stephan Walter 3e6e70d8e8 Add enum llama_ftype, sync ggml_type to model files (#709) 2 gadi atpakaļ
  Georgi Gerganov c3ac702e5e ggml : add ggml_cont() + optimize ggml_cpy() for contiguous dst 2 gadi atpakaļ
  comex f963b63afa Rewrite loading code to try to satisfy everyone: 2 gadi atpakaļ
  unbounded 62cfc54f77 Add quantize-stats command for testing quantization (#728) 2 gadi atpakaļ
  Georgi Gerganov 986b6ce9f9 ggml, llama : avoid heavy V transpose + improvements (#775) 2 gadi atpakaļ
  Marian Cepok c0bb1d3ce2 ggml : change ne to int64_t (#626) 2 gadi atpakaļ
  Justine Tunney 6f23ba5ee2 Ensure --mlock works properly with mmap() support 2 gadi atpakaļ
  Slaren c03ae8dca1 Add mmap support for model files 2 gadi atpakaļ
  Stephan Walter c1f885067c ggml : introduce structs for the q4 data blocks (#356) 2 gadi atpakaļ
  comex 563cdc391d Support calling mlock() on loaded model data on Linux and macOS (#453) 2 gadi atpakaļ
  Stephan Walter 69c92298a9 Deduplicate q4 quantization functions (#383) 2 gadi atpakaļ
  Georgi Gerganov f5a77a629b Introduce C-style API (#370) 2 gadi atpakaļ
  hoangmit 6eac39ba95 Add RMS norm and use it (#187) 2 gadi atpakaļ

Jaunāki Vecāki