github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 08:30:05 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[PR #7082] [MERGED] llama: fix compiler flag differences #12303

New Issue
Closed
opened 2026-04-12 23:54:51 -05:00 by GiteaMirror · 0 comments
GiteaMirror commented 2026-04-12 23:54:51 -05:00
Owner
Copy Link

📋 Pull Request Information

Original PR: https://github.com/ollama/ollama/pull/7082
Author: @dhiltgen
Created: 10/2/2024
Status: ✅ Merged
Merged: 10/3/2024
Merged by: @dhiltgen

Base: jmorganca/llama ← Head: go_server_flags

📝 Commits (1)

  • 61b5518 llama: fix compiler flag differences

📊 Changes

3 files changed (+19 additions, -15 deletions)

View changed files

📝 llama/llama.go (+8 -8)
📝 llama/make/Makefile.rocm (+6 -2)
📝 llama/make/cuda.make (+5 -5)

📄 Description

Adjust the flags for the new Go server to more closely match the generate flow

Still more refinement to do, but I think this gets us closer. Comparisons on an Ryzen 9 7950X system (windows) (linux)

This PR: llama3.2 tps

  • cpu: 9.17(w) 4.07(l)
  • cpu_avx: 14.19(w) 14.78(l)
  • cpu_avx2: 14.3(w) 15.3(l)

C++ server (0.3.12 release)

  • cpu: 5.9(w) 7.34(l)
  • cpu_avx: 17.5(w) 21.47(l)
  • cpu_avx2: 19 tps(w) 22.63(l)

🔄 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/7082 **Author:** [@dhiltgen](https://github.com/dhiltgen) **Created:** 10/2/2024 **Status:** ✅ Merged **Merged:** 10/3/2024 **Merged by:** [@dhiltgen](https://github.com/dhiltgen) **Base:** `jmorganca/llama` ← **Head:** `go_server_flags` --- ### 📝 Commits (1) - [`61b5518`](https://github.com/ollama/ollama/commit/61b55189446d7d4964a9867316428d8d79523b25) llama: fix compiler flag differences ### 📊 Changes **3 files changed** (+19 additions, -15 deletions) <details> <summary>View changed files</summary> 📝 `llama/llama.go` (+8 -8) 📝 `llama/make/Makefile.rocm` (+6 -2) 📝 `llama/make/cuda.make` (+5 -5) </details> ### 📄 Description Adjust the flags for the new Go server to more closely match the generate flow Still more refinement to do, but I think this gets us closer. Comparisons on an Ryzen 9 7950X system (windows) (linux) This PR: llama3.2 tps - cpu: 9.17(w) 4.07(l) - cpu_avx: 14.19(w) 14.78(l) - cpu_avx2: 14.3(w) 15.3(l) C++ server (0.3.12 release) - cpu: 5.9(w) 7.34(l) - cpu_avx: 17.5(w) 21.47(l) - cpu_avx2: 19 tps(w) 22.63(l) --- <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 23:54:51 -05:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label pull-request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#12303
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?