OpenNMT / OpenNMT/CTranslate2

Loading model on low CPU memory

Open
#1,411 4 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

enhancement
Dominant language
C++
Stars
4.7k
Forks
536
Avg merge
12h 12m
Merged PRs (30d)
4

Description

I am struggling to load a quantized model lacking sufficient CPU memory to load the weights.

Usually I would split the weights up in multiple shards and then load them accordingly.

Is this, or something similar, also possible in CTranslate?

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by examining CTranslate2's model-loading path and how quantized model weights are currently handled. The issue does not name files or tests, so first map the relevant loading code and existing memory behavior. Done means establishing whether sharded weight loading, or a comparable low-CPU-memory approach, is supported.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
machine-learning, performance
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.