GGUF layer package for running MiniMax-M3-UD-Q4_K_XL across a local Mesh LLM cluster.
This package is derived from unsloth/MiniMax-M3-GGUF and keeps the original GGUF distribution split into per-layer artifacts for distributed inference.
Highlights
| Run locally | Pool multiple machines | OpenAI-compatible | Package variant |
|---|---|---|---|
| Private inference on your hardware | Split layers across peers | Serve /v1/chat/completions locally | UD-Q4_K_XL layer package |
Model Overview
| Property | Value |
|---|---|
| Source model | unsloth/MiniMax-M3-GGUF |
| Model id | unsloth/MiniMax-M3-GGUF:UD-Q4_K_XL |
| Family | MiniMax |
| Parameter scale | not recorded |
| Quantization | UD-Q4_K_XL |
| Layer count | 60 |
| Activation width | 6144 |
| Package size | 247.2 GB |
| Source file | UD-Q4_K_XL/MiniMax-M3-UD-Q4_K_XL-00001-of-00007.gguf |
| Package repo | meshllm/MiniMax-M3-UD-Q4_K_XL-layers |
Recommended Use
- Local and private inference with Mesh LLM.
- Multi-machine serving when the full GGUF is too large for one host.
- OpenAI-compatible chat/completions workflows through Mesh LLM's local API.
For upstream architecture details, chat template guidance, sampling recommendations, license terms, and benchmark notes, see the source model card: unsloth/MiniMax-M3-GGUF.
Quickstart
# Run this on each machine that should contribute memory/compute.
mesh-llm serve --model "meshllm/MiniMax-M3-UD-Q4_K_XL-layers" --split
# Check the mesh and discover the OpenAI-compatible model name.
curl -s http://localhost:3131/api/status
curl -s http://localhost:3131/v1/models
# Send an OpenAI-compatible chat request.
curl -s http://localhost:3131/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "unsloth/MiniMax-M3-GGUF:UD-Q4_K_XL",
"messages": [{"role": "user", "content": "Write a tiny hello-world function in Rust."}],
"max_tokens": 128
}'
Package Variant
| Property | Value |
|---|---|
| Format | layer-package |
| Canonical source ref | unsloth/MiniMax-M3-GGUF@main/UD-Q4_K_XL/MiniMax-M3-UD-Q4_K_XL-00001-of-00007.gguf |
| Source revision | main |
| Source SHA-256 | e8e16c72310bd397e88e3ae5dc12d68dba4bf4d10e74825f91a9558e88f622a7 |
| Skippy ABI | 0.1.25 |
| Package manifest SHA-256 | d2b681ca6fa30a0610622aaf36ea1f6ce7aae47640aedb29b8431d983ef49e18 |
What Is Included
| Artifact | Path | Contents | SHA-256 |
|---|---|---|---|
| Manifest | model-package.json | Package schema, source identity, checksums | d2b681ca6fa30a0610622aaf36ea1f6ce7aae47640aedb29b8431d983ef49e18 |
| Metadata | shared/metadata.gguf | 0 tensors, 7.9 MB | 45bf2ed724916a3a54e74d159d0b21f56f311be85556f6b337facc50866840b6 |
| Embeddings | shared/embeddings.gguf | 1 tensors, 1.2 GB | 4313427784d01824f54727719c8381136c37c814324de700ec5c27a99d3f952b |
| Output head | shared/output.gguf | 2 tensors, 1.2 GB | 142a218816ca9132df37139747659ce2a5dcf756e128e112856b718b4ab67d66 |
| Transformer layers | layers/layer-*.gguf | 60 layer artifacts, 945 tensors, 244.8 GB | see model-package.json |
Validation
Generated by the Mesh LLM HF Jobs splitter from mesh-llm ref main.
Each artifact is checksummed as it is written, uploaded to this repository, and removed from the job workspace before the next artifact is produced.
skippy-model-package write-package "/source/UD-Q4_K_XL/MiniMax-M3-UD-Q4_K_XL-00001-of-00007.gguf" --out-dir "/tmp/meshllm-layer-job-meshllm_MiniMax-M3-UD-Q4_K_XL-layers-199/package"
Links
- Source model: unsloth/MiniMax-M3-GGUF
- Mesh LLM website: meshllm.cloud
- Mesh LLM: github.com/Mesh-LLM/mesh-llm
- Discord: discord.gg/rs6fmc63eN
- Package catalog: meshllm/catalog
- Package format: layer-package-repos.md