[PR #3465] [MERGED] fix metal gpu #42412

Closed
opened 2026-04-24 22:11:13 -05:00 by GiteaMirror · 0 comments
Owner

📋 Pull Request Information

Original PR: https://github.com/ollama/ollama/pull/3465
Author: @mxyng
Created: 4/2/2024
Status: Merged
Merged: 4/2/2024
Merged by: @mxyng

Base: mainHead: mxyng/fix-metal


📝 Commits (1)

📊 Changes

1 file changed (+2 additions, -2 deletions)

View changed files

📝 llm/server.go (+2 -2)

📄 Description

this includes two fixes for macos

  1. in the rare case a model's minimal memory does not fit into gpu, it will fallback to cpu which doesn't exist for metal devices
  2. num_gpu 0 should be passed to the subprocess since it's a completely valid configuration

🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.

## 📋 Pull Request Information **Original PR:** https://github.com/ollama/ollama/pull/3465 **Author:** [@mxyng](https://github.com/mxyng) **Created:** 4/2/2024 **Status:** ✅ Merged **Merged:** 4/2/2024 **Merged by:** [@mxyng](https://github.com/mxyng) **Base:** `main` ← **Head:** `mxyng/fix-metal` --- ### 📝 Commits (1) - [`80163eb`](https://github.com/ollama/ollama/commit/80163ebcb5f79661ae0c7094cc2983c87dab2e28) fix metal gpu ### 📊 Changes **1 file changed** (+2 additions, -2 deletions) <details> <summary>View changed files</summary> 📝 `llm/server.go` (+2 -2) </details> ### 📄 Description this includes two fixes for macos 1. in the rare case a model's minimal memory does not fit into gpu, it will fallback to `cpu` which doesn't exist for metal devices 2. num_gpu 0 should be passed to the subprocess since it's a completely valid configuration --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>
GiteaMirror added the pull-request label 2026-04-24 22:11:13 -05:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#42412