Commits

Breadcrumbs

History for

onmaster

User selector

All users

Datepicker

All time

Commit History

Commits on Sep 10, 2023

• convert: remove most of the n_mult usage in convert.py (ggerganov#3098)

  Green-Sky

  authored
  6eeb4d9

  Copy full SHA for 6eeb4d9

Commits on Sep 7, 2023

• convert : fix F32 ftype not being saved (ggerganov#3048)

  cebtenzzre

  authored
  6336d83

  Copy full SHA for 6336d83

Commits on Sep 5, 2023

• convert: fix convert.py not working with int filename_stem (ggerganov#3028)

  Show description for c9c3220
  

  Green-Sky

  and

  cebtenzzre

  authored
  c9c3220

  Copy full SHA for c9c3220

Commits on Sep 3, 2023

• convert.py : BPE fixes (ggerganov#2938)

  Show description for cff7b0b
  KerfuffleV2

  authored
  cff7b0b

  Copy full SHA for cff7b0b

Commits on Sep 1, 2023

• convert : fix another python 3.8 issue (ggerganov#2949)

  cebtenzzre

  authored
  bce1fef

  Copy full SHA for bce1fef

Commits on Aug 31, 2023

• scripts: Use local gguf package when running from repo (ggerganov#2927)

  Show description for aeefac4
  KerfuffleV2

  authored
  aeefac4

  Copy full SHA for aeefac4

• convert : fix python 3.8 support, modernize type annotations (ggerganov#2916)

  Show description for 92d0b75
  cebtenzzre

  authored
  92d0b75

  Copy full SHA for 92d0b75

Commits on Aug 30, 2023

• convert.py : use dir name to name the llama

  ggerganov

  committed
  b532a69

  Copy full SHA for b532a69

• convert : various script cleanups/fixes + merges and special token handling (ggerganov#2842)

  Show description for dc07dc4
  KerfuffleV2

  authored
  dc07dc4

  Copy full SHA for dc07dc4

Commits on Aug 29, 2023

• convert.py : fix baichuan7B support (ggerganov#2870)

  Show description for bcce96b
  

  jameswu2014

  and

  ggerganov

  authored
  bcce96b

  Copy full SHA for bcce96b

Commits on Aug 26, 2023

• convert.py : advanced option (ggerganov#2753)

  Show description for 730d9c6
  KerfuffleV2

  authored
  730d9c6

  Copy full SHA for 730d9c6

• Handle null rope scaling value (ggerganov#2793)

  pnb

  authored
  a2ca4e9

  Copy full SHA for a2ca4e9

Commits on Aug 25, 2023

• convert.py : Get rope scale from HuggingFace models (ggerganov#2772)

  Show description for 28b2c99
  pnb

  authored
  28b2c99

  Copy full SHA for 28b2c99

• convert.py : export rope freq_base when converting CodeLlama from an HF model (ggerganov#2773)

  slaren

  authored
  12e2e33

  Copy full SHA for 12e2e33

Commits on Aug 24, 2023

• convert.py : try to determine n_ctx automatically for CodeLlama (ggerganov#2770)

  slaren

  authored
  d0f77b1

  Copy full SHA for d0f77b1

• gguf : add rope_freq_base parameter for CodeLlama (ggerganov#2769)

  slaren

  authored
  0d3094f

  Copy full SHA for 0d3094f

• convert : auto-determine model name based on dir + scripts update

  ggerganov

  committed
  8f8c28e

  Copy full SHA for 8f8c28e

• fix convert.py for codellama, add llama 34B to the list of recognized models (ggerganov#2768)

  slaren

  authored
  fea95c6

  Copy full SHA for fea95c6

Commits on Aug 23, 2023

• llm : add Falcon support (ggerganov#2717)

  Show description for cf658ad
  

  3 peopleauthored
  cf658ad

  Copy full SHA for cf658ad

• chmod : make scripts executable (ggerganov#2675)

  cebtenzzre

  authored
  7c2227a

  Copy full SHA for 7c2227a

Commits on Aug 22, 2023

• convert.py : clarifying error message (ggerganov#2718)

  apetenchea

  authored
  3b6cfe7

  Copy full SHA for 3b6cfe7

• gguf : add ftype meta info to the model (ggerganov#2710)

  Show description for deb7dfc
  ggerganov

  authored
  deb7dfc

  Copy full SHA for deb7dfc

Commits on Aug 21, 2023

• gguf : new file format with flexible meta data (beta) (ggerganov#2398)

  Show description for 6381d4e
  

  6 peopleauthored
  6381d4e

  Copy full SHA for 6381d4e

Commits on Aug 6, 2023

• convert.py : add missing abstract methods for quantized data (ggerganov#2491)

  oss-dev-somewhere

  authored
  2e8265a

  Copy full SHA for 2e8265a

Commits on Jul 27, 2023

• convert.py : Update to support 70B HF format model files (ggerganov#2427)

  Show description for 7c529ce
  mj-shifu

  authored
  7c529ce

  Copy full SHA for 7c529ce

Commits on Jul 25, 2023

• convert.py : support bpe tokenizer (ggerganov#2228)

  Show description for fce48ca
  

  ftgreat

  and

  ldwang

  authored
  fce48ca

  Copy full SHA for fce48ca

Commits on Jul 23, 2023

• llama : grouped-query attention + LLaMAv2 70B support (ggerganov#2276)

  Show description for e76d630
  

  ggerganov

  and

  JohannesGaessler

  authored
  e76d630

  Copy full SHA for e76d630

Commits on Jul 19, 2023

• cmake : install targets (ggerganov#2256)

  Show description for b1f4290
  Freed-Wu

  authored
  b1f4290

  Copy full SHA for b1f4290

Commits on Jul 7, 2023

• convert.py: add mapping for safetensors bf16 (ggerganov#1598)

  Show description for 3e08ae9
  akx

  authored
  3e08ae9

  Copy full SHA for 3e08ae9

Commits on Jul 6, 2023

• convert : update for baichuan (ggerganov#2081)

  Show description for 36680f6
  

  foldl

  and

  Judd

  authored
  36680f6

  Copy full SHA for 36680f6

Commits on Jul 1, 2023

• convert : add support of baichuan-7b (ggerganov#2055)

  Show description for 471aab6
  

  foldl

  and

  Judd

  authored
  471aab6

  Copy full SHA for 471aab6

Commits on Jun 24, 2023

• convert : fix invalid params in write_vocab_only (ggerganov#1975)

  aisk

  authored
  c943d82

  Copy full SHA for c943d82

Commits on Jun 22, 2023

• rework convert.py to read hyper-parameters from config.json (ggerganov#1958)

  Show description for 7487137
  Green-Sky

  authored
  7487137

  Copy full SHA for 7487137

Commits on Jun 17, 2023

• hooks : setting up flake8 and pre-commit hooks (ggerganov#1681)

  Show description for 5ddf7ea
  jpodivin

  authored
  5ddf7ea

  Copy full SHA for 5ddf7ea

Commits on May 16, 2023

• convert.py: Support models which are stored in a single pytorch_model.bin (ggerganov#1469)

  Show description for 2b26469
  TheBloke

  authored
  2b26469

  Copy full SHA for 2b26469

Pagination

Previous

Next