Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Jul 14, 2024

• llama : fix Gemma-2 Query scaling factors (#8473)

  Show description for 73cf442
  

  ggerganov

  and

  danielhanchen

  authored
  73cf442

  Copy full SHA for 73cf442

• gguf_hash.py: Add sha256 (#8470)

  Show description for e236528
  

  mofosyne

  and

  compilade

  authored
  e236528

  Copy full SHA for e236528

• llama : fix pre-tokenization of non-special added tokens (#8228)

  Show description for fa79495
  compilade

  authored
  fa79495

  Copy full SHA for fa79495

Commits on Jul 13, 2024

• vulkan : cmake integration (#8119)

  Show description for 17eb6aa
  bandoti

  authored
  17eb6aa

  Copy full SHA for 17eb6aa

• metal : template-ify some of the kernels (#8447)

  Show description for c917b67
  ggerganov

  authored
  c917b67

  Copy full SHA for c917b67

Commits on Jul 12, 2024

• server : handle content array in chat API (#8449)

  Show description for 4e24cff
  

  ggerganov

  and

  ngxson

  authored
  4e24cff

  Copy full SHA for 4e24cff

• main : print error on empty input (#8456)

  ggerganov

  authored
  6af51c0

  Copy full SHA for 6af51c0

• llama : suppress unary minus operator warning (#8448)

  Show description for f532262
  danbev

  authored
  f532262

  Copy full SHA for f532262

• server : ensure batches are either all embed or all completion (#8420)

  Show description for c3ebcfa
  iamlemec

  authored
  c3ebcfa

  Copy full SHA for c3ebcfa

• docker : fix filename for convert-hf-to-gguf.py in tools.sh (#8441)

  Show description for 8a4441e
  kriation

  authored
  8a4441e

  Copy full SHA for 8a4441e

• convert : remove fsep token from GPTRefactForCausalLM (#8237)

  Show description for 5aefbce
  jpodivin

  authored
  5aefbce

  Copy full SHA for 5aefbce

• examples : sprintf -> snprintf (#8434)

  Show description for 71c1121
  ggerganov

  authored
  71c1121

  Copy full SHA for 71c1121

• ggml : minor naming changes (#8433)

  Show description for 370b1f7
  ggerganov

  authored
  370b1f7

  Copy full SHA for 370b1f7

• [SYCL] fix the mul_mat_id ut issues (#8427)

  Show description for b549a1b
  

  3 peopleauthored
  b549a1b

  Copy full SHA for b549a1b

Commits on Jul 11, 2024

• ggml : add NVPL BLAS support (#8329) (#8425)

  Show description for 3686456
  

  nicholaiTukanov

  and

  ntukanov

  authored
  3686456

  Copy full SHA for 3686456

• cuda : suppress 'noreturn' warn in no_device_code (#8414)

  Show description for b078c61
  danbev

  authored
  b078c61

  Copy full SHA for b078c61

• CUDA: optimize and refactor MMQ (#8416)

  Show description for 808aba3
  JohannesGaessler

  authored
  808aba3

  Copy full SHA for 808aba3

• gitignore : deprecated binaries

  ggerganov

  committed
  a977c11

  Copy full SHA for a977c11

• tokenize : add --no-parse-special option (#8423)

  Show description for 9a55ffe
  compilade

  authored
  9a55ffe

  Copy full SHA for 9a55ffe

• llama : use F32 precision in Qwen2 attention and no FA (#8412)

  ggerganov

  authored
  7a221b6

  Copy full SHA for 7a221b6

• Initialize default slot sampling parameters from the global context. (#8418)

  HanClinto

  authored
  278d0e1

  Copy full SHA for 278d0e1

Commits on Jul 10, 2024

• Name Migration: Build the deprecation-warning 'main' binary every time (#8404)

  Show description for dd07a12
  HanClinto

  authored
  dd07a12

  Copy full SHA for dd07a12

• [SYCL] Use multi_ptr to clean up deprecated warnings (#8256)

  

  AidanBeltonS

  authored
  f4444d9

  Copy full SHA for f4444d9

• ggml : move sgemm sources to llamafile subfolder (#8394)

  Show description for 6b2a849
  ggerganov

  authored
  6b2a849

  Copy full SHA for 6b2a849

• ggml : add AArch64 optimized GEMV and GEMM Q4 kernels (#5780)

  Show description for 0f1a39f
  Dibakar

  authored
  0f1a39f

  Copy full SHA for 0f1a39f

• gguf-py rel pipeline (#8410)

  Show description for 83321c6
  monatis

  authored
  83321c6

  Copy full SHA for 83321c6

• llama : C++20 compatibility for u8 strings (#8408)

  iboB

  authored
  cc61948

  Copy full SHA for cc61948

• msvc : silence codecvt c++17 deprecation warnings (#8395)

  iboB

  authored
  7a80710

  Copy full SHA for 7a80710

• llama : add assert about missing llama_encode() call (#8400)

  Show description for a8be1e6
  

  fairydreaming

  and

  sszymczy

  authored
  a8be1e6

  Copy full SHA for a8be1e6

• py : fix converter for internlm2 (#8321)

  Show description for e4dd31f
  RunningLeon

  authored
  e4dd31f

  Copy full SHA for e4dd31f

• py : fix extra space in convert_hf_to_gguf.py (#8407)

  laik

  authored
  8f0fad4

  Copy full SHA for 8f0fad4

Commits on Jul 9, 2024

• Server: Enable setting default sampling parameters via command-line (#8402)

  Show description for a59f8fd
  HanClinto

  authored
  a59f8fd

  Copy full SHA for a59f8fd

• Update README.md to fix broken link to docs (#8399)

  Show description for fd560fe
  andysalerno

  authored
  fd560fe

  Copy full SHA for fd560fe

• Deprecation warning to assist with migration to new binary names (#8283)

  Show description for e500d61
  HanClinto

  authored
  e500d61

  Copy full SHA for e500d61

• make/cmake: LLAMA_NO_CCACHE -> GGML_NO_CCACHE (#8392)

  JohannesGaessler

  authored
  a03e8dd

  Copy full SHA for a03e8dd

Pagination

Previous

Next