[PR #4110] [CLOSED] split binaries into metadata and data layers #16656

Closed
opened 2026-04-16 05:38:42 -05:00 by GiteaMirror · 0 comments
Owner

📋 Pull Request Information

Original PR: https://github.com/ollama/ollama/pull/4110
Author: @mxyng
Created: 5/3/2024
Status: Closed

Base: mainHead: mxyng/split-bin


📝 Commits (9)

📊 Changes

16 files changed (+811 additions, -577 deletions)

View changed files

📝 convert/convert.go (+2 -1)
📝 convert/gemma.go (+2 -13)
📝 convert/llama.go (+3 -10)
📝 convert/mistral.go (+2 -13)
📝 convert/mixtral.go (+3 -14)
📝 integration/utils_test.go (+1 -1)
llm/filetype.go (+140 -0)
📝 llm/ggla.go (+12 -0)
📝 llm/ggml.go (+19 -77)
📝 llm/gguf.go (+9 -2)
📝 llm/llm.go (+4 -52)
📝 server/images.go (+302 -321)
📝 server/layer.go (+32 -43)
server/model.go (+259 -0)
📝 server/routes.go (+1 -6)
📝 server/routes_test.go (+20 -24)

📄 Description

No description provided


🔄 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/4110 **Author:** [@mxyng](https://github.com/mxyng) **Created:** 5/3/2024 **Status:** ❌ Closed **Base:** `main` ← **Head:** `mxyng/split-bin` --- ### 📝 Commits (9) - [`1d8c850`](https://github.com/ollama/ollama/commit/1d8c850f38a444779f54187a0f764438280b740b) quantize any fp16/fp32 model - [`1909e62`](https://github.com/ollama/ollama/commit/1909e624cea1ba2ca0d1c20f0c216d395dc917e5) update tests - [`da0bb5d`](https://github.com/ollama/ollama/commit/da0bb5d77202ea546fe2428d2863f13b69ea4d6b) comments - [`fd071ea`](https://github.com/ollama/ollama/commit/fd071eab8bae682983774cf852fe2c7afb0993df) rebase - [`b535afe`](https://github.com/ollama/ollama/commit/b535afe35c4ea96c6b2bc2f429e31a5331053852) no iterator - [`3244a25`](https://github.com/ollama/ollama/commit/3244a25c7988538b19c52642a6a86a24080e7ce3) only quantize language models - [`122b35c`](https://github.com/ollama/ollama/commit/122b35c7840b96fdacb74e616f6816151c5aa01e) s/DisplayLongest/String/ - [`41ae232`](https://github.com/ollama/ollama/commit/41ae232e106aa6e43ae4ee72471cb3f6d1c97c56) split model layer into metadata and data layers - [`dc474f9`](https://github.com/ollama/ollama/commit/dc474f9b83c3a7fb9624c3cf38ce46a807127ca5) handle intermediate blobs ### 📊 Changes **16 files changed** (+811 additions, -577 deletions) <details> <summary>View changed files</summary> 📝 `convert/convert.go` (+2 -1) 📝 `convert/gemma.go` (+2 -13) 📝 `convert/llama.go` (+3 -10) 📝 `convert/mistral.go` (+2 -13) 📝 `convert/mixtral.go` (+3 -14) 📝 `integration/utils_test.go` (+1 -1) ➕ `llm/filetype.go` (+140 -0) 📝 `llm/ggla.go` (+12 -0) 📝 `llm/ggml.go` (+19 -77) 📝 `llm/gguf.go` (+9 -2) 📝 `llm/llm.go` (+4 -52) 📝 `server/images.go` (+302 -321) 📝 `server/layer.go` (+32 -43) ➕ `server/model.go` (+259 -0) 📝 `server/routes.go` (+1 -6) 📝 `server/routes_test.go` (+20 -24) </details> ### 📄 Description _No description provided_ --- <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 2026-04-16 05:38:42 -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#16656