Commits

Breadcrumbs

History for

onmaster

User selector

All users

Datepicker

All time

Commit History

Commits on Apr 9, 2024

• llama : add Command R Plus support (ggerganov#6491)

  Show description for 5dc9dd7
  

  5 peopleauthored
  5dc9dd7

Commits on Mar 26, 2024

• IQ1_M: 1.75 bpw quantization (ggerganov#6302)

  Show description for 55c1b2a
  

  ikawrakow

  and

  Kawrakow

  authored
  55c1b2a

Commits on Mar 12, 2024

• ggml : reuse quantum structs across backends (ggerganov#5943)

  Show description for 8030da7
  ggerganov

  authored
  8030da7

Commits on Mar 11, 2024

• Better 1.5 bit quantization (ggerganov#5971)

  Show description for be858f6
  

  ikawrakow

  and

  Kawrakow

  authored
  be858f6

Commits on Mar 9, 2024

• ggml : remove old quantization functions (ggerganov#5942)

  Show description for 5b09797
  ggerganov

  authored
  5b09797

• ggml : add ggml-common.h to deduplicate shared code (ggerganov#5940)

  Show description for 8a3012a
  ggerganov

  authored
  8a3012a

Commits on Feb 28, 2024

• ggml : make i-quants work with super-blocks of 64 (CPU,Metal) (ggerganov#5760)

  Show description for 7c4263d
  

  ikawrakow

  and

  Kawrakow

  authored
  7c4263d

Commits on Feb 27, 2024

• IQ4_XS: a 4.25 bpw quantization (ggerganov#5747)

  Show description for 0becb22
  

  ikawrakow

  and

  Kawrakow

  authored
  0becb22

Commits on Feb 26, 2024

• Adding IQ2_S and IQ2_M to complete coverage of the 2-3 bit quantization range (ggerganov#5721)

  Show description for a33e6a0
  

  3 peopleauthored
  a33e6a0

Commits on Feb 24, 2024

• IQ3_S: a much better alternative to Q3_K (ggerganov#5676)

  Show description for 4c4cb30
  

  ikawrakow

  and

  Kawrakow

  authored
  4c4cb30

Commits on Feb 21, 2024

• IQ4_NL: 4-bit non-linear quants with blocks of 32 (ggerganov#5590)

  Show description for a14679c
  

  ikawrakow

  and

  Kawrakow

  authored
  a14679c

Commits on Feb 18, 2024

• 1.5 bit quantization (ggerganov#5453)

  Show description for bd2d4e3
  

  ikawrakow

  and

  Kawrakow

  authored
  bd2d4e3

Commits on Feb 11, 2024

• ggml : add mmla kernels for quantized GEMM (ggerganov#4966)

  Show description for a07d0fe
  snadampal

  authored
  a07d0fe

Commits on Feb 5, 2024

• ggml : make use of ggml-quants.h possible in C++ code (ggerganov#5338)

  Show description for c6b3955
  

  ikawrakow

  and

  Kawrakow

  authored
  c6b3955

Commits on Jan 30, 2024

• SOTA 3-bit quants (ggerganov#5196)

  Show description for f4d7e54
  

  ikawrakow

  and

  Kawrakow

  authored
  f4d7e54

Commits on Jan 17, 2024

• ggml : add IQ2 to test-backend-ops + refactoring (ggerganov#4990)

  Show description for 3856668
  ggerganov

  authored
  3856668

Commits on Jan 16, 2024

• ggml : importance matrix support for legacy quants (ggerganov#4969)

  Show description for 334a835
  

  ikawrakow

  and

  Kawrakow

  authored
  334a835

Commits on Jan 14, 2024

• Add ability to use importance matrix for all k-quants (ggerganov#4930)

  Show description for 467a882
  

  ikawrakow

  and

  Kawrakow

  authored
  467a882

• 2-bit quantizations (ggerganov#4897)

  Show description for 147b17a
  

  ikawrakow

  and

  Kawrakow

  authored
  147b17a

Commits on Jan 11, 2024

• ggml : SOTA 2-bit quants (add IQ2_XS) (ggerganov#4856)

  Show description for 49662cb
  

  ikawrakow

  and

  Kawrakow

  authored
  49662cb

Commits on Jan 8, 2024

• SOTA 2-bit quants (ggerganov#4773)

  Show description for dd5ae06
  

  ikawrakow

  and

  Kawrakow

  authored
  dd5ae06

Commits on Jan 5, 2024

• ggml : fix q2_k bpw in comments (ggml/680)

  ggerganov

  committed
  d061bf9

Commits on Oct 30, 2023

• ggml : move FP16 <-> FP32 code to ggml-impl.h (ggerganov#3861)

  Show description for 207b519
  ggerganov

  authored
  207b519

Commits on Oct 29, 2023

• ggml : quantization refactoring (ggerganov#3833)

  Show description for d69d777
  

  ggerganov

  and

  cebtenzzre

  authored
  d69d777

Loading