Commits

Branch selector

 master

User selector

KerfuffleV2

Datepicker

All time

Commit History

Commits on Dec 5, 2023

• llama : allow overriding GGUF metadata when loading model (#4092)

  Show description for 5aa365d
  

  3 peopleauthored
  5aa365d

  Copy full SHA for 5aa365d

Commits on Nov 18, 2023

• tokenize example: Respect normal add BOS token behavior (#4126)

  Show description for 28a2e6e
  KerfuffleV2

  authored
  28a2e6e

  Copy full SHA for 28a2e6e

• Clean up ggml-cuda.cu warnings when compiling with clang (for ROCM) (#4124)

  Show description for 2923f17
  KerfuffleV2

  authored
  2923f17

  Copy full SHA for 2923f17

Commits on Nov 17, 2023

• Respect tokenizer.ggml.add_bos_token value when tokenizing (#4040)

  Show description for 91f6499
  KerfuffleV2

  authored
  91f6499

  Copy full SHA for 91f6499

Commits on Nov 13, 2023

• Add ReLU and SQR CUDA ops to (partially) fix Persimmon offloading (#4041)

  Show description for bb50a79
  KerfuffleV2

  authored
  bb50a79

  Copy full SHA for bb50a79

Commits on Nov 12, 2023

• gguf-py: gguf_writer: Use bytearray to build metadata (#4051)

  Show description for 21fd874
  KerfuffleV2

  authored
  21fd874

  Copy full SHA for 21fd874

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 5, 2023

• Allow common process_escapes to handle \x sequences (#3928)

  Show description for d9ccce2
  KerfuffleV2

  authored
  d9ccce2

  Copy full SHA for d9ccce2

Commits on Nov 4, 2023

• gguf-py: Support 01.AI Yi models (#3943)

  KerfuffleV2

  authored
  f28af0d

  Copy full SHA for f28af0d

Commits on Nov 2, 2023

• cuda : add ROCM aliases for CUDA pool stuff (#3918)

  KerfuffleV2

  authored
  629f917

  Copy full SHA for 629f917

Commits on Oct 29, 2023

• Extend llama_kv_cache_seq_rm to allow matching any sequence (#3843)

  Show description for 6e08281
  KerfuffleV2

  authored
  6e08281

  Copy full SHA for 6e08281

Commits on Oct 28, 2023

• llama : allow quantizing k-quants to fall back when tensor size incompatible (#3747)

  Show description for bd6d9e2
  KerfuffleV2

  authored
  bd6d9e2

  Copy full SHA for bd6d9e2

Commits on Oct 27, 2023

• speculative : ensure draft and target model vocab matches (#3812)

  Show description for 41aee4d
  KerfuffleV2

  authored
  41aee4d

  Copy full SHA for 41aee4d

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 11, 2023

• common : fix mirostat state when using multiple sequences (#3543)

  Show description for 70c29da
  KerfuffleV2

  authored
  70c29da

  Copy full SHA for 70c29da

Commits on Oct 8, 2023

• llama : fix missing break in Persimmon arch case statements (#3535)

  KerfuffleV2

  authored
  63d3b06

  Copy full SHA for 63d3b06

Commits on Oct 7, 2023

• Fix trying to strip newline from empty prompt and cfg prompt file content (#3534)

  KerfuffleV2

  authored
  a16e89c

  Copy full SHA for a16e89c

Commits on Oct 6, 2023

• kv cache slot search improvements (#3493)

  Show description for 9ca79d5
  KerfuffleV2

  authored
  9ca79d5

  Copy full SHA for 9ca79d5

Commits on Oct 4, 2023

• convert : fix Baichuan2 models by using vocab size in config.json (#3299)

  Show description for 019ba1d
  KerfuffleV2

  authored
  019ba1d

  Copy full SHA for 019ba1d

Commits on Sep 14, 2023

• gguf-py : support identity operation in TensorNameMap (#3095)

  Show description for e394084
  KerfuffleV2

  authored
  e394084

  Copy full SHA for e394084

Commits on Sep 6, 2023

• convert-llama-ggml-to-gguf: Try to handle files older than GGJTv3 (#3023)

  Show description for ea2c85d
  KerfuffleV2

  authored
  ea2c85d

  Copy full SHA for ea2c85d

Commits on Sep 3, 2023

• gguf(python): Fix special vocab handling when id < 0 (#2984)

  KerfuffleV2

  authored
  6519e9c

  Copy full SHA for 6519e9c

• convert.py : BPE fixes (#2938)

  Show description for cff7b0b
  KerfuffleV2

  authored
  cff7b0b

  Copy full SHA for cff7b0b

Commits on Sep 2, 2023

• logging: Fix creating empty file even when disabled (#2966)

  Show description for 3358c38
  

  KerfuffleV2

  and

  staviq

  authored
  3358c38

  Copy full SHA for 3358c38

Commits on Sep 1, 2023

• Allow quantize to only copy tensors, some other improvements (#2931)

  Show description for 5d6f19f
  KerfuffleV2

  authored
  5d6f19f

  Copy full SHA for 5d6f19f

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

Commits on Aug 30, 2023

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

  Show description for dc07dc4
  KerfuffleV2

  authored
  dc07dc4

  Copy full SHA for dc07dc4

Commits on Aug 26, 2023

• convert.py : advanced option (#2753)

  Show description for 730d9c6
  KerfuffleV2

  authored
  730d9c6

  Copy full SHA for 730d9c6

Commits on Aug 24, 2023

• Fix for main example getting stuck when -n -2 and --interactive (#2767)

  Show description for 7694add
  KerfuffleV2

  authored
  7694add

  Copy full SHA for 7694add

Commits on Aug 22, 2023

• Improve handling of special tokens in GGML to GGUF converter (#2725)

  Show description for 777f42b
  KerfuffleV2

  authored
  777f42b

  Copy full SHA for 777f42b

• Fix convert-llama-ggmlv3-to-gguf.py vocab conversion (#2698)

  Show description for 930523c
  KerfuffleV2

  authored
  930523c

  Copy full SHA for 930523c

Commits on Aug 17, 2023

• Add --cfg-negative-prompt-file option for examples (#2591)

  Show description for 8dae7ce
  KerfuffleV2

  authored
  8dae7ce

  Copy full SHA for 8dae7ce

Commits on Jun 17, 2023

• Allow cmake to build ggml as a library (#1896)

  Show description for b4c6f46
  KerfuffleV2

  authored
  b4c6f46

  Copy full SHA for b4c6f46

Commits on Jun 13, 2023

• Allow "quantizing" to f16 and f32 (#1787)

  Show description for 74d4cfa
  KerfuffleV2

  authored
  74d4cfa

  Copy full SHA for 74d4cfa

Commits on Jun 11, 2023

• Fix issue where interactive mode crashes when input exceeds ctx size (#1789)

  Show description for fa84c4b
  KerfuffleV2

  authored
  fa84c4b

  Copy full SHA for fa84c4b

Pagination

Previous

Next