[PR #3567] [MERGED] partial offloading #11200

Closed
opened 2026-04-12 23:24:00 -05:00 by GiteaMirror · 0 comments
Owner

📋 Pull Request Information

Original PR: https://github.com/ollama/ollama/pull/3567
Author: @mxyng
Created: 4/10/2024
Status: Merged
Merged: 4/10/2024
Merged by: @mxyng

Base: mxyng/tensor-layerHead: mxyng/mem


📝 Commits (1)

📊 Changes

6 files changed (+96 additions, -84 deletions)

View changed files

📝 format/bytes.go (+1 -1)
📝 gpu/gpu.go (+3 -3)
📝 gpu/gpu_darwin.go (+3 -4)
📝 gpu/types.go (+1 -1)
📝 llm/ggml.go (+42 -35)
📝 llm/server.go (+46 -40)

📄 Description

partial offloading of a model requires more memory in most cases (except for metal)


🔄 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/3567 **Author:** [@mxyng](https://github.com/mxyng) **Created:** 4/10/2024 **Status:** ✅ Merged **Merged:** 4/10/2024 **Merged by:** [@mxyng](https://github.com/mxyng) **Base:** `mxyng/tensor-layer` ← **Head:** `mxyng/mem` --- ### 📝 Commits (1) - [`b2201b8`](https://github.com/ollama/ollama/commit/b2201b87cc5f9937fcddb2a8163da701a8764799) partial offloading ### 📊 Changes **6 files changed** (+96 additions, -84 deletions) <details> <summary>View changed files</summary> 📝 `format/bytes.go` (+1 -1) 📝 `gpu/gpu.go` (+3 -3) 📝 `gpu/gpu_darwin.go` (+3 -4) 📝 `gpu/types.go` (+1 -1) 📝 `llm/ggml.go` (+42 -35) 📝 `llm/server.go` (+46 -40) </details> ### 📄 Description partial offloading of a model requires more memory in most cases (except for metal) --- <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 2026-04-12 23:24:00 -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#11200