Commits

Commit history

1. Commits on Dec 27, 2025

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

     JohannesGaessler

     authored
     a52dc60View commit details

     Copy full SHA for a52dc60

     View code at this pointBrowse repository at this point
2. 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
     b1f3a6eView commit details

     Copy full SHA for b1f3a6e

     View code at this pointBrowse repository at this point
3. Commits on Nov 25, 2025

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

     taronaeo

     authored
     877566dView commit details

     Copy full SHA for 877566d

     View code at this pointBrowse repository at this point
4. Commits on Nov 7, 2025

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

     Show description for 9008027
     CISC

     authored
     9008027View commit details

     Copy full SHA for 9008027

     View code at this pointBrowse repository at this point

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

     Show description for 16bcc12
     ggerganov

     authored
     16bcc12View commit details

     Copy full SHA for 16bcc12

     View code at this pointBrowse repository at this point
5. Commits on Nov 2, 2025

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

     Show description for cd5e3b5
     ggerganov

     authored
     cd5e3b5View commit details

     Copy full SHA for cd5e3b5

     View code at this pointBrowse repository at this point

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

     Show description for 76af40a
     alundb

     authored
     76af40aView commit details

     Copy full SHA for 76af40a

     View code at this pointBrowse repository at this point
6. Commits on Oct 30, 2025

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

     Show description for d261223
     

     7 peopleauthored
     d261223View commit details

     Copy full SHA for d261223

     View code at this pointBrowse repository at this point
7. Commits on Oct 6, 2025

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

     Show description for 3df2244
     

     Gadflyii

     and

     slaren

     authored
     3df2244View commit details

     Copy full SHA for 3df2244

     View code at this pointBrowse repository at this point
8. Commits on Oct 3, 2025

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

     Show description for f6dcda3
     

     ddh0

     and

     ggerganov

     authored
     f6dcda3View commit details

     Copy full SHA for f6dcda3

     View code at this pointBrowse repository at this point
9. Commits on Sep 24, 2025

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

     Show description for e789095
     JohannesGaessler

     authored
     e789095View commit details

     Copy full SHA for e789095

     View code at this pointBrowse repository at this point
10. Commits on Sep 5, 2025

    • aLoRA Support (#15327)

      Show description for fd62188
      

      gabe-l-hart

      and

      CISC

      authored
      fd62188View commit details

      Copy full SHA for fd62188

      View code at this pointBrowse repository at this point
11. Commits on Aug 31, 2025

    • sampling : optimize samplers by reusing bucket sort (#15665)

      Show description for e92d53b
      

      ggerganov

      and

      JohannesGaessler

      authored
      e92d53bView commit details

      Copy full SHA for e92d53b

      View code at this pointBrowse repository at this point
12. Commits on Aug 30, 2025

    • llama: use FA + max. GPU layers by default (#15434)

      Show description for e81b8e4
      JohannesGaessler

      authored
      e81b8e4View commit details

      Copy full SHA for e81b8e4

      View code at this pointBrowse repository at this point
13. Commits on Aug 28, 2025

    • model : jina-embeddings-v3 support (#13693)

      Show description for 84ab83c
      CISC

      authored
      84ab83cView commit details

      Copy full SHA for 84ab83c

      View code at this pointBrowse repository at this point
14. Commits on Aug 22, 2025

    • llama : remove KV cache defragmentation logic (#15473)

      Show description for 9ebebef
      ggerganov

      authored
      9ebebefView commit details

      Copy full SHA for 9ebebef

      View code at this pointBrowse repository at this point
15. Commits on Aug 21, 2025

    • llama : remove deprecated llama_kv_self API (#15472)

      Show description for cd36b5e
      ggerganov

      authored
      cd36b5eView commit details

      Copy full SHA for cd36b5e

      View code at this pointBrowse repository at this point

    • kv-cache : drop the "unified" prefix (#15467)

      Show description for 715a6db
      ggerganov

      authored
      715a6dbView commit details

      Copy full SHA for 715a6db

      View code at this pointBrowse repository at this point
16. Commits on Aug 14, 2025

    • server : add SWA checkpoints (#15293)

      Show description for d32e03f
      ggerganov

      authored
      d32e03fView commit details

      Copy full SHA for d32e03f

      View code at this pointBrowse repository at this point

    • finetune: SGD optimizer, more CLI args (#13873)

      Show description for 5cdb27e
      

      3 peopleauthored
      5cdb27eView commit details

      Copy full SHA for 5cdb27e

      View code at this pointBrowse repository at this point
17. Commits on Aug 5, 2025

    • llama : add gpt-oss (#15091)

      Show description for fd1234c
      

      3 peopleauthored
      fd1234cView commit details

      Copy full SHA for fd1234c

      View code at this pointBrowse repository at this point
18. Commits on Jul 31, 2025

    • llama : allow other bufts when overriding to CPU, add --no-repack option (#14990)

      slaren

      authored
      d6818d0View commit details

      Copy full SHA for d6818d0

      View code at this pointBrowse repository at this point

    • Add LLaDA 8b Diffusion model (#14771)

      Show description for 8a4a856
      am17an

      authored
      8a4a856View commit details

      Copy full SHA for 8a4a856

      View code at this pointBrowse repository at this point
19. Commits on Jul 24, 2025

    • context : perform output reorder lazily upon access after sync (#14853)

      Show description for e4868d1
      ggerganov

      authored
      e4868d1View commit details

      Copy full SHA for e4868d1

      View code at this pointBrowse repository at this point
20. Commits on Jul 17, 2025

    • llama : reuse compute graphs (#14482)

      Show description for 01612b7
      ggerganov

      authored
      01612b7View commit details

      Copy full SHA for 01612b7

      View code at this pointBrowse repository at this point
21. Commits on Jul 16, 2025

    • llama : add high-throughput mode (#14363)

      Show description for 225e7a1
      

      ggerganov

      and

      JohannesGaessler

      authored
      225e7a1View commit details

      Copy full SHA for 225e7a1

      View code at this pointBrowse repository at this point

    • Support diffusion models: Add Dream 7B (#14644)

      Show description for ab14019
      am17an

      authored
      ab14019View commit details

      Copy full SHA for ab14019

      View code at this pointBrowse repository at this point

    • llama: add LLAMA_API to deprecated llama_kv_self_seq_div (#14708)

      Show description for 79e0b68
      

      Min-Hua

      and

      Min-Hua Chen

      authored
      79e0b68View commit details

      Copy full SHA for 79e0b68

      View code at this pointBrowse repository at this point
22. Commits on Jul 15, 2025

    • model : add PLaMo-2 support (#14560)

      Show description for 68e37a6
      

      4 peopleauthored
      68e37a6View commit details

      Copy full SHA for 68e37a6

      View code at this pointBrowse repository at this point
23. Commits on Jul 11, 2025

    • llama : move enum llama_vocab_pre_type to implementation (#14631)

      Show description for 0d5375d
      ggerganov

      authored
      0d5375dView commit details

      Copy full SHA for 0d5375d

      View code at this pointBrowse repository at this point
24. Commits on Jul 8, 2025

    • model : add hunyuan moe (#14425)

      Show description for 8f22dc0
      

      ngxson

      and

      kooshi

      authored
      8f22dc0View commit details

      Copy full SHA for 8f22dc0

      View code at this pointBrowse repository at this point
25. Commits on Jun 23, 2025

    • kv-cells : fix tracking of seq_pos (#14339)

      Show description for 7b50d58
      ggerganov

      authored
      7b50d58View commit details

      Copy full SHA for 7b50d58

      View code at this pointBrowse repository at this point
26. Commits on Jun 22, 2025

    • quantize : handle user-defined pruning of whole layers (blocks) (#13037)

      EAddario

      authored
      fa4a9f2View commit details

      Copy full SHA for fa4a9f2

      View code at this pointBrowse repository at this point
27. Commits on Jun 20, 2025

    • vocab : prevent tokenizer overflow (#14301)

      Show description for dd6e6d0
      retr0reg

      authored
      dd6e6d0View commit details

      Copy full SHA for dd6e6d0

      View code at this pointBrowse repository at this point

    • llama : improve sep token handling (#14272)

      CISC

      authored
      88fc854View commit details

      Copy full SHA for 88fc854

      View code at this pointBrowse repository at this point