Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

MPT support in llama.cpp #3417

Merged
merged 17 commits into from Oct 10, 2023
Merged

MPT support in llama.cpp #3417

merged 17 commits into from Oct 10, 2023

Commits on Sep 30, 2023

  1. CUDA: added support for ggml_clamp (see also: ggerganov/ggml#545)

    @jploski
    jploski committed Sep 30, 2023
    Configuration menu
    Copy the full SHA
    b49792b View commit details
    Browse the repository at this point in the history

  2. mpt : added an implementation based (mostly) on falcon integration, m… 

    …odified with deltas from ggml/examples/mpt
    @jploski
    jploski committed Sep 30, 2023
    Configuration menu
    Copy the full SHA
    15236e8 View commit details
    Browse the repository at this point in the history

  3. mpt : protect against "clip_qkv": null in mpt-7b

    @jploski
    jploski committed Sep 30, 2023
    Configuration menu
    Copy the full SHA
    84e30e8 View commit details
    Browse the repository at this point in the history

  4. mpt : quick fix to avoid "Strange model" warning when quantizing MPT … 

    …models
    @jploski
    jploski committed Sep 30, 2023
    Configuration menu
    Copy the full SHA
    00e8c5c View commit details
    Browse the repository at this point in the history

  5. mpt : addendum to changeset:84e30e8 - leave parameter clamp_kqv out f… 

    …rom metadata rather than use 0.0 to indicate "no clamping" (more compliant with the current GGUF spec?)
    @jploski
    jploski committed Sep 30, 2023
    Configuration menu
    Copy the full SHA
    1be89c4 View commit details
    Browse the repository at this point in the history

  6. mpt : standardized all tensor names to follow GGUF spec

    @jploski
    jploski committed Sep 30, 2023
    Configuration menu
    Copy the full SHA
    26c253e View commit details
    Browse the repository at this point in the history

  7. mpt : addendum to changeset:1be89c40 - use "req" parameter of GGUF_GE… 

    …T_KEY macro instead of duplicate code
    @jploski
    jploski committed Sep 30, 2023
    Configuration menu
    Copy the full SHA
    df072d2 View commit details
    Browse the repository at this point in the history

Commits on Oct 2, 2023

  1. mpt : fixed comment s/gptneox/mpt/

    @jploski
    jploski committed Oct 2, 2023
    Configuration menu
    Copy the full SHA
    90e7d6d View commit details
    Browse the repository at this point in the history

  2. mpt : remove tabs, trailing whitespace

    @jploski
    jploski committed Oct 2, 2023
    Configuration menu
    Copy the full SHA
    4708012 View commit details
    Browse the repository at this point in the history

Commits on Oct 3, 2023

  1. mpt : removed ne01 + n_past == ne00 assertion from alibi (cuda/f32) a… 

    …nd rope_shift from build_mpt
    @jploski
    jploski committed Oct 3, 2023
    Configuration menu
    Copy the full SHA
    1364bcd View commit details
    Browse the repository at this point in the history

Commits on Oct 9, 2023

  1. mpt : updated convert-mpt-hf-to-gguf.py to reflect changes made to co… 

    …nvert-gptneox-hf-to-gguf.py in pr:3252
    @jploski @cebtenzzre
    jploski authored and cebtenzzre committed Oct 9, 2023
    Configuration menu
    Copy the full SHA
    7d6a24a View commit details
    Browse the repository at this point in the history

  2. Merge branch 'master' of https://github.com/ggerganov/llama.cpp into … 

    …pull-3417
    @cebtenzzre
    cebtenzzre committed Oct 9, 2023
    Configuration menu
    Copy the full SHA
    292363e View commit details
    Browse the repository at this point in the history

  3. comment out n_past instead of marking it unused

    @cebtenzzre
    cebtenzzre committed Oct 9, 2023
    Configuration menu
    Copy the full SHA
    ad3c2f3 View commit details
    Browse the repository at this point in the history

  4. mpt : removed hardcoded +178 from convert script in favor of utilizin… 

    …g hparams["vocab_size"]
    @jploski
    jploski committed Oct 9, 2023
    Configuration menu
    Copy the full SHA
    1a454eb View commit details
    Browse the repository at this point in the history

  5. mpt : remove unused tokenizer_json in convert script

    @cebtenzzre
    cebtenzzre committed Oct 9, 2023
    Configuration menu
    Copy the full SHA
    32172f1 View commit details
    Browse the repository at this point in the history

Commits on Oct 10, 2023

  1. ggml : remove obsolete n_past assert in ggml_alibi

    @ggerganov
    ggerganov committed Oct 10, 2023
    Configuration menu
    Copy the full SHA
    96cf3f5 View commit details
    Browse the repository at this point in the history

  2. llama : print clam_kqv and max_alibi_bias hparams

    @ggerganov
    ggerganov committed Oct 10, 2023
    Configuration menu
    Copy the full SHA
    9b66378 View commit details
    Browse the repository at this point in the history