Историја ревизија

Аутор SHA1 Порука Датум
  Jesse Gross a1cda80bcb model: Update encoder cache to use multimodal input processing handler пре 2 месеци
  Jesse Gross 4614fafae0 ollamarunner: Don't panic for unimplemented features at runtime. пре 2 месеци
  Jesse Gross 0daaaef8c9 ollamarunner: Quiet debug logging and panic on unimplemented features пре 2 месеци
  Parth Sareen 0682dae027 sample: improve ollama engine sampler performance (#9374) пре 2 месеци
  Jesse Gross a7e63b82be ollamarunner: Improve multimodal input handling пре 2 месеци
  Jesse Gross b70fc4d51e model: Don't unconditionally add special tokens пре 2 месеци
  Michael Yang 05a01fdecb ml/backend/ggml: consolidate system info logging пре 2 месеци
  Jesse Gross 21aa666a1e ml: Enable support for flash attention пре 2 месеци
  Michael Yang 31e472baa4 runner: defer context cancel пре 2 месеци
  Bruce MacDonald 0c1041ad85 runner: default to greedy sampler for performance (#9407) пре 2 месеци
  Michael Yang d6af13efed runner: simplify tensor split parsing пре 2 месеци
  Parth Sareen 0b7e1676eb sample: add sampling package for new engine (#8410) пре 2 месеци
  Jesse Gross bd6a7d5e64 ollamarunner: Pass runner performance parameters to backends пре 3 месеци
  Daniel Hiltgen df2680b4b9 Wire up system info log for new engine (#9123) пре 3 месеци
  Jesse Gross ed443a0393 Runner for Ollama engine пре 5 месеци