Commits

Breadcrumbs

History for

on87038fe

User selector

All users

Datepicker

All time

Commit History

Commits on Feb 14, 2024

• llava : support v1.6 (#5267)

  Show description for aa23412
  

  3 peopleauthored
  aa23412

  Copy full SHA for aa23412

Commits on Feb 7, 2024

• convert : fix TypeError on GPT-2 vocab.json (#5288)

  

  Sang-Kil Park

  authored
  f68664a

  Copy full SHA for f68664a

Commits on Feb 6, 2024

• py : handle byte tokens in get_token_type (#5341)

  Show description for 906cff5
  ggerganov

  authored
  906cff5

  Copy full SHA for 906cff5

Commits on Jan 29, 2024

• py : fix except (#5194)

  Show description for 14fef85
  ggerganov

  authored
  14fef85

  Copy full SHA for 14fef85

• py : improve BPE tokenizer support (#5189)

  

  Sang-Kil Park

  authored
  e76627b

  Copy full SHA for e76627b

Commits on Jan 20, 2024

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

  cebtenzzre

  authored
  b43ebde

  Copy full SHA for b43ebde

Commits on Jan 18, 2024

• convert.py : fix llama/llama2 conversion due to vocab_size=-1 (#5019)

  Show description for b467577
  databyte

  authored
  b467577

  Copy full SHA for b467577

Commits on Jan 17, 2024

• py : fix whitespace

  ggerganov

  committed
  0f83e72

  Copy full SHA for 0f83e72

• py : fix missing added_tokens_dict for SPM and BPE vocabs (#4971)

  Show description for 4f4bf35
  ggerganov

  authored
  4f4bf35

  Copy full SHA for 4f4bf35

Commits on Jan 9, 2024

• convert.py : fix vanilla LLaMA model conversion (#4818)

  Show description for 6efb8eb
  

  teleprint-me

  and

  ggerganov

  authored
  6efb8eb

  Copy full SHA for 6efb8eb

Commits on Dec 27, 2023

• llama : add AWQ for llama, llama2, mpt, and mistral models (#4593)

  Show description for f679349
  

  4 peopleauthored
  f679349

  Copy full SHA for f679349

• Add byte token type when tokenizer.model is not exists (#4641)

  Show description for f56d607
  strutive07

  authored
  f56d607

  Copy full SHA for f56d607

Commits on Dec 14, 2023

• convert : support loading vocab from fast tokenizer config (#3633)

  Show description for 873637a
  

  strutive07

  and

  cebtenzzre

  authored
  873637a

  Copy full SHA for 873637a

Commits on Dec 13, 2023

• llama : add Mixtral support (#4406)

  Show description for 799a1cb
  

  3 peopleauthored
  799a1cb

  Copy full SHA for 799a1cb

Commits on Dec 12, 2023

• english : use typos to fix comments and logs (#4354)

  richardkiss

  authored
  9494d7c

  Copy full SHA for 9494d7c

Commits on Nov 30, 2023

• convert.py : fix llama/llama2 conversion due to vocab_size=-1 (#4258)

  slaren

  authored
  f4d973c

  Copy full SHA for f4d973c

Commits on Nov 25, 2023

• Update docs for yarn_ext_factor <0.0 as unspecified instead of NaN (#4189)

  crasm

  authored
  3014b54

  Copy full SHA for 3014b54

Commits on Nov 20, 2023

• ci : add flake8 to github actions (python linting) (#4129)

  Show description for f23c035
  Galunid

  authored
  f23c035

  Copy full SHA for f23c035

Commits on Nov 17, 2023

• convert : use 'model' value if it exists. This allows karpathy/tinyllamas to load (#4089)

  Show description for 2ab0707
  dmahurin

  authored
  2ab0707

  Copy full SHA for 2ab0707

Commits on Nov 14, 2023

• convert.py: also look for plain model.safetensors (#4043)

  Show description for b46d12f
  afrideva

  authored
  b46d12f

  Copy full SHA for b46d12f

Commits on Nov 11, 2023

• gguf-py: Refactor and allow reading/modifying existing GGUF files (#3981)

  Show description for 34b0a08
  

  KerfuffleV2

  and

  cebtenzzre

  authored
  34b0a08

  Copy full SHA for 34b0a08

Commits on Nov 9, 2023

• scripts: Generalize convert scripts (#3838)

  Show description for a75fa57
  Galunid

  authored
  a75fa57

  Copy full SHA for a75fa57

Commits on Nov 1, 2023

• llama : implement YaRN RoPE scaling (#2268)

  Show description for 898aeca
  

  cebtenzzre

  and

  jquesnelle

  authored
  898aeca

  Copy full SHA for 898aeca

Commits on Oct 28, 2023

• convert : ignore tokens if their IDs are within [0, vocab_size) (#3831)

  ggerganov

  authored
  8a2f2fe

  Copy full SHA for 8a2f2fe

Commits on Oct 22, 2023

• llama : validate special token ids are in range when loading GGUF model (#3635)

  Show description for a5e7dbd
  

  KerfuffleV2

  and

  ggerganov

  authored
  a5e7dbd

  Copy full SHA for a5e7dbd

Commits on Oct 20, 2023

• gguf : support big endian platform (#3552)

  Show description for 8cf19d6
  

  chenqiny

  and

  ggerganov

  authored
  8cf19d6

  Copy full SHA for 8cf19d6

Commits on Oct 3, 2023

• Work on the BPE tokenizer (#3252)

  Show description for ff5a3f0
  

  goerch

  and

  ggerganov

  authored
  ff5a3f0

  Copy full SHA for ff5a3f0

Commits on Oct 2, 2023

• gguf : general usability improvements (#3409)

  cebtenzzre

  authored
  0fe3210

  Copy full SHA for 0fe3210

Commits on Sep 27, 2023

• convert : remove bug in convert.py permute function (#3364)

  jzhang38

  authored
  e519621

  Copy full SHA for e519621

Commits on Sep 10, 2023

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

  Green-Sky

  authored
  6eeb4d9

  Copy full SHA for 6eeb4d9

Commits on Sep 7, 2023

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

  cebtenzzre

  authored
  6336d83

  Copy full SHA for 6336d83

Commits on Sep 5, 2023

• convert: fix convert.py not working with int filename_stem (#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 (#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 (#2949)

  cebtenzzre

  authored
  bce1fef

  Copy full SHA for bce1fef

Commits on Aug 31, 2023

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

  Show description for aeefac4
  KerfuffleV2

  authored
  aeefac4

  Copy full SHA for aeefac4

Pagination

Previous

Next