Commits

Breadcrumbs

History for

onmaster

User selector

All users

All time

Commit History

Commits on Aug 2, 2023

• tests : Fix compilation warnings (Linux/GCC) (#2451)

  Show description for 81844fb
  netrunnereve

  authored
  81844fb

  Copy full SHA for 81844fb

Commits on Jul 24, 2023

• make rms_norm_eps a parameter (#2374)

  Show description for 41c6741
  slaren

  authored
  41c6741

  Copy full SHA for 41c6741

• ggml : sync (unary ops refactor, static-correctness) (#2370)

  Show description for 5b2b2dc
  ggerganov

  authored
  5b2b2dc

  Copy full SHA for 5b2b2dc

Commits on Jul 19, 2023

• cmake : install targets (#2256)

  Show description for b1f4290
  Freed-Wu

  authored
  b1f4290

  Copy full SHA for b1f4290

Commits on Jul 18, 2023

• ci : integrate with ggml-org/ci (#2250)

  Show description for d01bccd
  ggerganov

  authored
  d01bccd

  Copy full SHA for d01bccd

Commits on Jul 11, 2023

• ggml : sync (abort callback, mul / add broadcast, fix alibi) (#2183)

  ggerganov

  authored
  20d7740

  Copy full SHA for 20d7740

Commits on Jul 10, 2023

• mpi : add support for distributed inference via MPI (#2099)

  Show description for 5656d10
  

  evanmiller

  and

  ggerganov

  authored
  5656d10

  Copy full SHA for 5656d10

Commits on Jul 7, 2023

• ggml : change ggml_graph_compute() API to not require context (#1999)

  Show description for 1d656d6
  

  mqy

  and

  ggerganov

  authored
  1d656d6

  Copy full SHA for 1d656d6

Commits on Jul 5, 2023

• tests : fix test-grad0

  ggerganov

  committed
  1b6efea

  Copy full SHA for 1b6efea

• ggml : generalize `quantize_fns` for simpler FP16 handling (#1237)

  Show description for 1b107b8
  

  sw

  and

  ggerganov

  authored
  1b107b8

  Copy full SHA for 1b107b8

Commits on Jun 26, 2023

• tests : fix quantize perf (#1990)

  Show description for a84ab1d
  katsu560

  authored
  a84ab1d

  Copy full SHA for a84ab1d

Commits on Jun 24, 2023

• tests : sync test-grad0 from ggml

  ggerganov

  committed
  65bdd52

  Copy full SHA for 65bdd52

• llama : fix top-p sampling to match the canonical definition (#1953)

  Show description for b061ba9
  alexrenda

  authored
  b061ba9

  Copy full SHA for b061ba9

• llama : make model stateless and context stateful (llama_state) (#1797)

  Show description for 527b6fb
  

  didzis

  and

  ggerganov

  authored
  527b6fb

  Copy full SHA for 527b6fb

Commits on Jun 16, 2023

• build : fix and ignore MSVC warnings (#1889)

  iboB

  authored
  9cbf50c

  Copy full SHA for 9cbf50c

Commits on Jun 13, 2023

• train : improved training-from-scratch example (#1652)

  Show description for e32089b
  

  xaedes

  and

  ggerganov

  authored
  e32089b

  Copy full SHA for e32089b

Commits on Jun 5, 2023

• ggml : add SOTA 2,3,4,5,6 bit k-quantizations (#1684)

  Show description for 99009e7
  

  3 peopleauthored
  99009e7

  Copy full SHA for 99009e7

Commits on May 19, 2023

• tests : add missing header

  ggerganov

  committed
  6986c78

  Copy full SHA for 6986c78

• minor : fix compile warnings

  ggerganov

  committed
  4b7e245

  Copy full SHA for 4b7e245

Commits on May 13, 2023

• ggml : implement backward pass for llama + small training-llama-from-scratch example (#1360)

  Show description for f954edd
  

  xaedes

  and

  ggerganov

  authored
  f954edd

  Copy full SHA for f954edd

Commits on May 6, 2023

• Remove default arguments from sampling functions (#1343)

  j-f1

  authored
  3924088

  Copy full SHA for 3924088

Commits on May 2, 2023

• llama : fix compile warnings

  ggerganov

  committed
  0e6cbff

  Copy full SHA for 0e6cbff

Commits on Apr 29, 2023

• llama : new sampling algorithms (#1126)

  Show description for dd7eff5
  ivanstepanovftw

  authored
  dd7eff5

  Copy full SHA for dd7eff5

Commits on Apr 25, 2023

• ggml : add Q8_0 quantization format (rename the old one to Q8_1) (ARM NEON) (#1179)

  Show description for 7a32fcb
  ggerganov

  authored
  7a32fcb

  Copy full SHA for 7a32fcb

Commits on Apr 22, 2023

• Fix CI: ARM NEON, quantization unit tests, editorconfig (#1122)

  sw

  authored
  c50b628

  Copy full SHA for c50b628

• ggml : unit test for quantization functions (#953)

  Show description for 5f93949
  unbounded

  authored
  5f93949

  Copy full SHA for 5f93949

Commits on Apr 17, 2023

• llama : well-defined static initialization of complex objects (#927)

  Show description for efd0564
  arikpoz

  authored
  efd0564

  Copy full SHA for efd0564

Commits on Mar 28, 2023

• tests : free llama context at the end of the test

  ggerganov

  committed
  d502bc7

  Copy full SHA for d502bc7

• all : be more strict about converting float to double (#458)

  Show description for 436e561
  

  sw

  and

  ggerganov

  authored
  436e561

  Copy full SHA for 436e561

• ggml : introduce structs for the q4 data blocks (#356)

  Show description for c1f8850
  

  sw

  and

  ggerganov

  authored
  c1f8850

  Copy full SHA for c1f8850

Commits on Mar 25, 2023

• Overhaul the examples structure

  Show description for a316a42
  ggerganov

  committed
  a316a42

  Copy full SHA for a316a42

Commits on Mar 22, 2023

• Deduplicate q4 quantization functions (#383)

  Show description for 69c9229
  

  sw

  and

  ggerganov

  authored
  69c9229

  Copy full SHA for 69c9229

• Introduce C-style API (#370)

  Show description for f5a77a6
  ggerganov

  authored
  f5a77a6

  Copy full SHA for f5a77a6

Commits on Mar 21, 2023

• Add tokenizer test + revert to C++11 (#355)

  Show description for eb34620
  ggerganov

  authored
  eb34620

  Copy full SHA for eb34620

Loading