Commits

Breadcrumbs

History for

onmaster

User selector

All users

All time

Commit History

Commits on Apr 6, 2026

• ggml: add Q1_0 1-bit quantization support (CPU) (#21273)

  Show description for 2e1f0a8
  

  khosravipasha

  and

  CISC

  authored
  2e1f0a8

  Copy full SHA for 2e1f0a8

Commits on Apr 1, 2026

• llama : refactor llama_model_quantize_params to expose a pure C interface (#20346)

  Show description for 4951250
  

  EAddario

  and

  ggerganov

  authored
  4951250

  Copy full SHA for 4951250

Commits on Mar 25, 2026

• llama: fix llama-model-saver (#20503)

  Show description for 36dafba
  

  JohannesGaessler

  and

  Siddhesh2377

  authored
  36dafba

  Copy full SHA for 36dafba

Commits on Mar 18, 2026

• llama : re-enable manual LoRA adapter free (#19983)

  Show description for 312cf03
  PopFlamingo

  authored
  312cf03

  Copy full SHA for 312cf03

Commits on Mar 11, 2026

• ggml : add NVFP4 quantization type support (#19769)

  Show description for 5eae9cb
  

  richarddd

  and

  CISC

  authored
  5eae9cb

  Copy full SHA for 5eae9cb

Commits on Mar 8, 2026

• llama: end-to-end tests (#19802)

  Show description for a976ff0
  JohannesGaessler

  authored
  a976ff0

  Copy full SHA for a976ff0

Commits on Mar 5, 2026

• chore : correct typos [no ci] (#20041)

  Show description for 92f7da0
  

  3 peopleauthored
  92f7da0

  Copy full SHA for 92f7da0

Commits on Feb 20, 2026

• quantize : add --dry-run option (#19526)

  Show description for 492bc31
  

  ddh0

  and

  CISC

  authored
  492bc31

  Copy full SHA for 492bc31

Commits on Feb 14, 2026

• llama : update LoRA API. + fix excessive graph reserves (#19280)

  Show description for 2d8015e
  

  3 peopleauthored
  2d8015e

  Copy full SHA for 2d8015e

Commits on Feb 12, 2026

• llama : update outdated comment in llama.h (#19428)

  Show description for f488429
  

  MonkeybreadSoftware

  and

  danbev

  authored
  f488429

  Copy full SHA for f488429

Commits on Feb 11, 2026

• llama : correct typos 'occured' and 'occurences' (#19414)

  Show description for 8ee538c
  thecaptain789

  authored
  8ee538c

  Copy full SHA for 8ee538c

Commits on Jan 28, 2026

• llama : disable Direct IO by default (#19109)

  Show description for c5c64f7
  ggerganov

  authored
  c5c64f7

  Copy full SHA for c5c64f7

Commits on Jan 25, 2026

• llama: fix integer type consistency in split helpers (#18894)

  Show description for 24bc238
  MaheshJakkala

  authored
  24bc238

  Copy full SHA for 24bc238

Commits on Jan 24, 2026

• llama-fit-params: keep explicit --ctx-size 0 (#19070)

  JohannesGaessler

  authored
  e9fd8dc

  Copy full SHA for e9fd8dc

Commits on Jan 15, 2026

• llama : add adaptive-p sampler (#17927)

  Show description for 13f1e4a
  ddh0

  authored
  13f1e4a

  Copy full SHA for 13f1e4a

• context : reserve new scheduler when graph topology changes (#18547)

  Show description for 39173bc
  ggerganov

  authored
  39173bc

  Copy full SHA for 39173bc

• lora: make sure model keep track of associated adapters (#18490)

  Show description for a7e6ddb
  

  ngxson

  and

  ggerganov

  authored
  a7e6ddb

  Copy full SHA for a7e6ddb

Commits on Jan 9, 2026

• server : use different seeds for child completions (#18700)

  Show description for f5f8812
  ggerganov

  authored
  f5f8812

  Copy full SHA for f5f8812

Commits on Jan 8, 2026

• llama-fit-params: free memory target per device (#18679)

  JohannesGaessler

  authored
  64848de

  Copy full SHA for 64848de

• llama : add `use_direct_io` flag for model loading (#18166)

  Show description for 2038101
  

  3 peopleauthored
  2038101

  Copy full SHA for 2038101

Commits on Jan 5, 2026

• model : add LFM2-ColBert-350M (#18607)

  Show description for 73d284a
  tdakhran

  authored
  73d284a

  Copy full SHA for 73d284a

Commits on Jan 4, 2026

• sampling : add support for backend sampling (#17004)

  Show description for d3dce4e
  

  4 peopleauthored
  d3dce4e

  Copy full SHA for d3dce4e

Commits on Dec 30, 2025

• lora: count lora nodes in graph_max_nodes (#18469)

  Show description for cd78e57
  ngxson

  authored
  cd78e57

  Copy full SHA for cd78e57

Commits on Dec 27, 2025

• llama: fix magic number of 999 for GPU layers (#18266)

  Show description for 026d2ad
  JohannesGaessler

  authored
  026d2ad

  Copy full SHA for 026d2ad

• llama_fit_params: return enum for fail vs. error (#18374)

  JohannesGaessler

  authored
  a52dc60

  Copy full SHA for a52dc60

Commits on Dec 15, 2025

• llama: automatically set parameters not set by the user in such a way that maximizes GPU utilization (#16653)

  Show description for b1f3a6e
  JohannesGaessler

  authored
  b1f3a6e

  Copy full SHA for b1f3a6e

Commits on Nov 25, 2025

• llama: introduce support for model-embedded sampling parameters (#17120)

  taronaeo

  authored
  877566d

  Copy full SHA for 877566d

Commits on Nov 7, 2025

• hparams : add n_embd_inp() to support extended embed (#16928)

  Show description for 9008027
  CISC

  authored
  9008027

  Copy full SHA for 9008027

• kv-cache : pad the cache size to 256 for performance (#17046)

  Show description for 16bcc12
  ggerganov

  authored
  16bcc12

  Copy full SHA for 16bcc12

Commits on Nov 2, 2025

• server : support unified cache across slots (#16736)

  Show description for cd5e3b5
  ggerganov

  authored
  cd5e3b5

  Copy full SHA for cd5e3b5

• docs: remove llama_sampler_accept reference in sampling sample usage (#16920)

  Show description for 76af40a
  alundb

  authored
  76af40a

  Copy full SHA for 76af40a

Commits on Oct 30, 2025

• model: add support for qwen3vl series (#16780)

  Show description for d261223
  

  7 peopleauthored
  d261223

  Copy full SHA for d261223

Commits on Oct 6, 2025

• llama : add --no-host to disable host buffers (#16310)

  Show description for 3df2244
  

  Gadflyii

  and

  slaren

  authored
  3df2244

  Copy full SHA for 3df2244

Commits on Oct 3, 2025

• server : context checkpointing for hybrid and recurrent models (#16382)

  Show description for f6dcda3
  

  ddh0

  and

  ggerganov

  authored
  f6dcda3

  Copy full SHA for f6dcda3

Commits on Sep 24, 2025

• llama: print memory breakdown on exit (#15860)

  Show description for e789095
  JohannesGaessler

  authored
  e789095

  Copy full SHA for e789095