Commits

Branch selector

 master

User selector

cebtenzzre

Datepicker

All time

Commit History

Commits on May 2, 2025

• llama-model : support Qwen2 embedding models and pooling_mode_lasttoken (#13245)

  cebtenzzre

  authored
  2f56761

  Copy full SHA for 2f56761

• convert : use correct context length for nomic-embed-text-v2 (#13216)

  cebtenzzre

  authored
  7d21234

  Copy full SHA for 7d21234

Commits on May 1, 2025

• llama-model : fix the reported size class for nomic-embed-text-v2-moe (#13223)

  cebtenzzre

  authored
  a70183e

  Copy full SHA for a70183e

Commits on Jun 10, 2024

• cmake : fix CMake requirement for CUDA (#7821)

  cebtenzzre

  authored
  864a99e

  Copy full SHA for 864a99e

Commits on May 16, 2024

• convert : get general.name from model dir, not its parent (#5615)

  Show description for dda64fc
  

  cebtenzzre

  and

  mofosyne

  authored
  dda64fc

  Copy full SHA for dda64fc

Commits on May 8, 2024

• cmake : fix typo (#7151)

  cebtenzzre

  authored
  4426e29

  Copy full SHA for 4426e29

Commits on Apr 9, 2024

• BERT tokenizer fixes (#6498)

  Show description for 1b67731
  cebtenzzre

  authored
  1b67731

  Copy full SHA for 1b67731

Commits on Mar 28, 2024

• convert : refactor vocab selection logic (#6355)

  cebtenzzre

  authored
  be55134

  Copy full SHA for be55134

Commits on Mar 26, 2024

• wpm : portable unicode tolower (#6305)

  Show description for 32c8486
  cebtenzzre

  authored
  32c8486

  Copy full SHA for 32c8486

Commits on Mar 23, 2024

• use _wfopen instead of fopen on Windows (#6248)

  Show description for 94d1b3b
  cebtenzzre

  authored
  94d1b3b

  Copy full SHA for 94d1b3b

Commits on Mar 20, 2024

• server tests : more pythonic process management; fix bare `except:` (#6146)

  Show description for bd60d82
  

  cebtenzzre

  and

  phymbert

  authored
  bd60d82

  Copy full SHA for bd60d82

Commits on Mar 18, 2024

• mpt : implement backwards compatiblity with duped output tensor (#6139)

  cebtenzzre

  authored
  d199ca7

  Copy full SHA for d199ca7

Commits on Mar 6, 2024

• ggml : use SYS_get_cpu if SYS_getcpu is not defined (#5906)

  Show description for e04e04f
  cebtenzzre

  authored
  e04e04f

  Copy full SHA for e04e04f

Commits on Mar 5, 2024

• quants : use MM256_SET_M128I consistently to fix gcc 7 build (#5889)

  cebtenzzre

  authored
  bd83694

  Copy full SHA for bd83694

Commits on Mar 2, 2024

• convert : automatically fall back to HfVocab if tokenizer.model doesn't exist (#5821)

  cebtenzzre

  authored
  4d4d236

  Copy full SHA for 4d4d236

• convert-hf : make model class definitions self-contained (#5825)

  cebtenzzre

  authored
  c7a0ad8

  Copy full SHA for c7a0ad8

Commits on Feb 23, 2024

• convert : fix missing ftype for gemma (#5690)

  cebtenzzre

  authored
  54fbcd2

  Copy full SHA for 54fbcd2

Commits on Feb 22, 2024

• mpt : do not duplicate token_embd.weight on disk (#5670)

  cebtenzzre

  authored
  15499eb

  Copy full SHA for 15499eb

Commits on Feb 21, 2024

• examples : do not assume BOS when shifting context (#5622)

  cebtenzzre

  authored
  89febfe

  Copy full SHA for 89febfe

Commits on Feb 19, 2024

• make : pass CPPFLAGS directly to nvcc, not via -Xcompiler (#5598)

  cebtenzzre

  authored
  f24ed14

  Copy full SHA for f24ed14

Commits on Feb 18, 2024

• build : pass all warning flags to nvcc via -Xcompiler (#5570)

  Show description for a0c2dad
  cebtenzzre

  authored
  a0c2dad

  Copy full SHA for a0c2dad

Commits on Feb 13, 2024

• llama : add support for Nomic Embed (#5468)

  cebtenzzre

  authored
  ea9c8e1

  Copy full SHA for ea9c8e1

Commits on Feb 3, 2024

• YaRN : store rope scaling type as int32_t in memory (#5285)

  Show description for 1ec3332
  cebtenzzre

  authored
  1ec3332

  Copy full SHA for 1ec3332

Commits on Jan 31, 2024

• kompute : llama-bench support and ggml_cpu_has_kompute() (#5226)

  cebtenzzre

  authored
  e8dc55d

  Copy full SHA for e8dc55d

Commits on Jan 29, 2024

• kompute : fix fallback to CPU (#5201)

  cebtenzzre

  authored
  6daa69e

  Copy full SHA for 6daa69e

• Nomic Vulkan backend (#4456)

  Show description for fbf1dde
  

  7 peopleauthored
  fbf1dde

  Copy full SHA for fbf1dde

Commits on Jan 26, 2024

• cmake : pass CPU architecture flags to nvcc (#5146)

  cebtenzzre

  authored
  bbe7c56

  Copy full SHA for bbe7c56

Commits on Jan 25, 2024

• examples : make pydantic scripts pass mypy and support py3.8 (#5099)

  cebtenzzre

  authored
  d292f4f

  Copy full SHA for d292f4f

Commits on Jan 20, 2024

• convert : partially revert PR #4818 (#5041)

  cebtenzzre

  authored
  b43ebde

  Copy full SHA for b43ebde

• perplexity : fix MSVC build after #5020 (#5043)

  Show description for 97c1549
  cebtenzzre

  authored
  97c1549

  Copy full SHA for 97c1549

Commits on Dec 21, 2023

• common : remove incorrect --model-draft default (#4568)

  cebtenzzre

  authored
  8fe03ff

  Copy full SHA for 8fe03ff

Commits on Dec 18, 2023

• decode : fix logits_valid for legacy API (#4516)

  cebtenzzre

  authored
  2994f0c

  Copy full SHA for 2994f0c

Commits on Dec 17, 2023

• gguf-py : fail fast on nonsensical special token IDs (#4489)

  cebtenzzre

  authored
  f7f468a

  Copy full SHA for f7f468a

Commits on Dec 16, 2023

• llama : sanity checks for access to logits (#4274)

  Show description for 8a5be3b
  

  cebtenzzre

  and

  ggerganov

  authored
  8a5be3b

  Copy full SHA for 8a5be3b

Commits on Dec 13, 2023

• build : detect host compiler and cuda compiler separately (#4414)

  cebtenzzre

  authored
  70f806b

  Copy full SHA for 70f806b

Pagination

Previous

Next