Commit History

Autor SHA1 Mensaxe Data
  Jesse Gross f66216e399 ggml: Support heterogeneous KV cache layer sizes in memory estimation hai 1 mes
  frob 7c168b08c9 server: add missing function parens to debug log (#9255) hai 2 meses
  Michael Yang 58245413f4 next ollama runner (#7913) hai 2 meses
  Stefan Weil abfdc4710f all: fix typos in documentation, code, and comments (#7021) hai 4 meses
  Jesse Gross 6cd566872b sched: Lift parallel restriction for multimodal models except mllama hai 6 meses
  Daniel Hiltgen 05cd82ef94 Rename gpu package discover (#7143) hai 6 meses
  Patrick Devine abed273de3 add "stop" command (#6739) hai 7 meses
  Daniel Hiltgen 90ca84172c Fix embeddings memory corruption (#6467) hai 8 meses
  Richard Lyons 885cf45087 Fix white space. hai 8 meses
  Richard Lyons 9352eeb752 Reset NumCtx. hai 8 meses
  Richard Lyons 0ad0e738cd Override numParallel only if unset. hai 8 meses
  Michael Yang 2697d7f5aa lint hai 8 meses
  Michael Yang b732beba6a lint hai 9 meses
  Michael Yang 5c1912769e Merge pull request #5473 from ollama/mxyng/environ hai 9 meses
  Daniel Hiltgen 345420998e Prevent partial loading on mixed GPU brands hai 9 meses
  Michael Yang 85d9d73a72 comments hai 9 meses
  Michael Yang 0f1910129f int hai 10 meses
  Michael Yang 8570c1c0ef keepalive hai 10 meses
  Michael Yang 55cd3ddcca bool hai 10 meses
  Jeffrey Morgan 791650ddef sched: only error when over-allocating system memory (#5626) hai 9 meses
  Jeffrey Morgan e4ff73297d server: fix model reloads when setting `OLLAMA_NUM_PARALLEL` (#5560) hai 9 meses
  Jeffrey Morgan 0ee87615c7 sched: don't error if paging to disk on Windows and macOS (#5523) hai 9 meses
  Daniel Hiltgen af28b94533 Merge pull request #5469 from dhiltgen/prevent_system_oom hai 10 meses
  Daniel Hiltgen 955f2a4e03 Only set default keep_alive on initial model load hai 10 meses
  Daniel Hiltgen 3c75113e37 Prevent loading models larger than total memory hai 10 meses
  Daniel Hiltgen cff3f44f4a Fix case for NumCtx hai 10 meses
  Daniel Hiltgen 3518aaef33 Merge pull request #4218 from dhiltgen/auto_parallel hai 10 meses
  Blake Mizerany cb42e607c5 llm: speed up gguf decoding by a lot (#5246) hai 10 meses
  Daniel Hiltgen 9929751cc8 Disable concurrency for AMD + Windows hai 10 meses
  Daniel Hiltgen 17b7186cd7 Enable concurrency by default hai 1 ano