Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Jul 9, 2024

• OpenAI v1/completions: allow stop token list (ollama#5551)

  Show description for 4918fae
  royjhan

  authored
  4918fae

  Copy full SHA for 4918fae

• separate request tests (ollama#5578)

  royjhan

  authored
  0aff678

  Copy full SHA for 0aff678

• Detect CUDA OS Overhead

  Show description for f6f759f
  dhiltgen

  committed
  f6f759f

  Copy full SHA for f6f759f

• Merge pull request ollama#5579 from dhiltgen/win_static_deps

  Show description for 9544a57
  dhiltgen

  authored
  9544a57

  Copy full SHA for 9544a57

• Statically link c++ and thread lib

  Show description for b51e3b6
  dhiltgen

  committed
  b51e3b6

  Copy full SHA for b51e3b6

• Merge pull request ollama#5440 from ollama/mxyng/messages-templates

  Show description for 6bbbc50
  mxyng

  authored
  6bbbc50

  Copy full SHA for 6bbbc50

• Merge pull request ollama#5126 from ollama/mxyng/messages

  Show description for 9bbddc3
  mxyng

  authored
  9bbddc3

  Copy full SHA for 9bbddc3

• server: fix model reloads when setting OLLAMA_NUM_PARALLEL (ollama#5560)

  Show description for e4ff732
  jmorganca

  authored
  e4ff732

  Copy full SHA for e4ff732

• Bundle missing CRT libraries

  Show description for b44320d
  dhiltgen

  committed
  b44320d

  Copy full SHA for b44320d

Commits on Jul 8, 2024

• Workaround broken ROCm p2p copy

  Show description for 0bacb30
  dhiltgen

  committed
  0bacb30

  Copy full SHA for 0bacb30

Commits on Jul 7, 2024

• llm: remove ambiguous comment when putting upper limit on predictions to avoid infinite generation (ollama#5535)

  jmorganca

  authored
  53da2c6

  Copy full SHA for 53da2c6

• llm: allow gemma 2 to context shift (ollama#5534)

  jmorganca

  authored
  d8def1f

  Copy full SHA for d8def1f

• Update llama.cpp submodule to a8db2a9c (ollama#5530)

  jmorganca

  authored
  571dc61

  Copy full SHA for 571dc61

• llm: print caching notices in debug only (ollama#5533)

  jmorganca

  authored
  0e09c38

  Copy full SHA for 0e09c38

• sched: don't error if paging to disk on Windows and macOS (ollama#5523)

  jmorganca

  authored
  0ee8761

  Copy full SHA for 0ee8761

Commits on Jul 6, 2024

• gpu: report system free memory instead of 0 (ollama#5521)

  jmorganca

  authored
  f8241bf

  Copy full SHA for f8241bf

• llm: add -DBUILD_SHARED_LIBS=off to common cpu cmake flags (ollama#5520)

  jmorganca

  authored
  4607c70

  Copy full SHA for 4607c70

• release: move mingw library cleanup to correct job

  jmorganca

  committed
  c12f1c5

  Copy full SHA for c12f1c5

• release: remove unwanted mingw dll.a files

  jmorganca

  committed
  a08f20d

  Copy full SHA for a08f20d

• Revert "llm: only statically link libstdc++"

  Show description for 6cea036
  jmorganca

  committed
  6cea036

  Copy full SHA for 6cea036

• llm: only statically link libstdc++

  jmorganca

  committed
  5796bfc

  Copy full SHA for 5796bfc

• llm: statically link pthread and stdc++ dependencies in windows build

  jmorganca

  committed
  f1a379a

  Copy full SHA for f1a379a

• llm: add GGML_STATIC flag to windows static lib

  jmorganca

  committed
  9ae1469

  Copy full SHA for 9ae1469

• llm: add COMMON_DARWIN_DEFS to arm static build (ollama#5513)

  jmorganca

  authored
  e0348d3

  Copy full SHA for e0348d3

• llm: fix missing dylibs by restoring old build behavior on Linux and macOS (ollama#5511)

  Show description for 2cc854f
  jmorganca

  authored
  2cc854f

  Copy full SHA for 2cc854f

Commits on Jul 5, 2024

• llm: put back old include dir (ollama#5507)

  Show description for 5304b76
  jmorganca

  authored
  5304b76

  Copy full SHA for 5304b76

• update named templates

  mxyng

  committed
  fb6cbc0

  Copy full SHA for fb6cbc0

• fix cmake build (ollama#5505)

  jmorganca

  authored
  4fd5f35

  Copy full SHA for 4fd5f35

• Merge pull request ollama#5502 from dhiltgen/ci_fixes

  Show description for 842f85f
  dhiltgen

  authored
  842f85f

  Copy full SHA for 842f85f

• Always go build in CI generate steps

  Show description for 9d30f9f
  dhiltgen

  committed
  9d30f9f

  Copy full SHA for 9d30f9f

• types/model: remove knowledge of digest (ollama#5500)

  Show description for 631cfd9
  bmizerany

  authored
  631cfd9

  Copy full SHA for 631cfd9

• no funcs

  mxyng

  committed
  326363b

  Copy full SHA for 326363b

• fix model reloading

  Show description for ac7a842
  mxyng

  committed
  ac7a842

  Copy full SHA for ac7a842

• comments

  mxyng

  committed
  2c3fe1f

  Copy full SHA for 2c3fe1f

• update message processing

  mxyng

  committed
  269ed6e

  Copy full SHA for 269ed6e

Pagination

Previous

Next