replicate/cog

Pick latest CUDA patch version when specifying minor version

Closed

#813 opened on Nov 27, 2022

View on GitHubContributor guide
 (2 comments) (0 reactions) (0 assignees)Python (356 forks)batch import
good first issuenexttype/bug

Repository metrics

Stars
 (5,724 stars)
PR merge metrics
 (Avg merge 7d 8h) (19 merged PRs in 30d)

Description

build:
  python_version: "3.10"
  cuda: "11.6"
  python_packages:
   - "torch==1.13.0"

This automatically selects CUDA 11.6.0 for the base image. It should pick the latest patch version and use 11.6.2.

Contributor guide

Research direction
Investigate how the cog build system resolves CUDA versions. Look at the Dockerfile generation code to find where '11.6' is mapped to a full version. Modify that part to query the latest patch version from NVIDIA's repository or use a package manager like apt to get the latest version. Ensure backward compatibility.
Tech stack
pythondocker
Domain
devops
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.
1-3 hours
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Stale
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
Basic PythonDocker usageCUDA versioning
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
50