ggml-org/llama.cpp

split: allow --split-max-size option

Open

#6,259 opened on Mar 23, 2024

View on GitHub
 (4 comments) (1 reaction) (0 assignees)C++ (18,202 forks)batch import
enhancementgood first issuehelp wantedsplit

Repository metrics

Stars
 (110,169 stars)
PR merge metrics
 (Avg merge 6d 8h) (389 merged PRs in 30d)

Description

Motivation

we support --split-max-tensors since:

  • #6135

As mentionned by @Artefact2 in this comment:

allowing to split by file size would be more intuitive (and usually more appropriate since file size is usually the limiting factor, eg 4G for FAT or 50G for HF)

Proposition:

Introduce --split-max-size N(M|G) split strategy to split files in file with a max size of N Megabytes or Gigabytes. As it is not possible to have less than 1 tensor per GGUF, this size is a soft limit.

Contributor guide

Research direction
Study the existing split implementation in llama.cpp, particularly the split max tensors option, and then implement a split max size option that limits file size by considering tensor sizes.
Tech stack
cpp
Domain
cli
Issue type
Feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
Half day
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
C++Git
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
65