GGUF layer package for running ERNIE-4.5-21B-A3B-PT-UD-Q4_K_XL across a local Mesh LLM cluster.
This package is derived from unsloth/ERNIE-4.5-21B-A3B-PT-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/ERNIE-4.5-21B-A3B-PT-GGUF |
| Model id | unsloth/ERNIE-4.5-21B-A3B-PT-GGUF:UD-Q4_K_XL |
| Family | ERNIE |
| Parameter scale | 21B-A3B |
| Quantization | UD-Q4_K_XL |
| Layer count | 28 |
| Activation width | 2560 |
| Package size | 12.6 GB |
| Source file | ERNIE-4.5-21B-A3B-PT-UD-Q4_K_XL.gguf |
| Package repo | meshllm/ERNIE-4.5-21B-A3B-PT-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/ERNIE-4.5-21B-A3B-PT-GGUF.
Quickstart
# Run this on each machine that should contribute memory/compute.
mesh-llm serve --model "meshllm/ERNIE-4.5-21B-A3B-PT-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/ERNIE-4.5-21B-A3B-PT-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/ERNIE-4.5-21B-A3B-PT-GGUF@main/ERNIE-4.5-21B-A3B-PT-UD-Q4_K_XL.gguf |
| Source revision | main |
| Source SHA-256 | 8aa12d9e7b0e8199fda963573a6d70fa6d357c5ce5217459c8f4723ddd03e6ff |
| Skippy ABI | 0.1.24 |
| Package manifest SHA-256 | 5e2820e2bae1ec30c6193b01e8c7eafff46b1fbaff2f6fedf1d82f564803feed |
What Is Included
| Artifact | Path | Contents | SHA-256 |
|---|---|---|---|
| Manifest | model-package.json | Package schema, source identity, checksums | 5e2820e2bae1ec30c6193b01e8c7eafff46b1fbaff2f6fedf1d82f564803feed |
| Metadata | shared/metadata.gguf | 0 tensors, 2.3 MB | 13596196817092a0c7805be5a7a1e0b4345400ba1e82a4a7addf50ff28791f55 |
| Embeddings | shared/embeddings.gguf | 1 tensors, 209.4 MB | 4cc1dc66b4be64fe1206ed78d43c00bff048cca5993227b1a990100e60199b12 |
| Output head | shared/output.gguf | 1 tensors, 2.3 MB | 1fde9970ff4af0994c7b7b411c0475c15721f65e1f1e2bed3c538f2b3a800348 |
| Transformer layers | layers/layer-*.gguf | 28 layer artifacts, 387 tensors, 12.4 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/ERNIE-4.5-21B-A3B-PT-UD-Q4_K_XL.gguf" --out-dir "/tmp/meshllm-layer-job-meshllm_ERNIE-4.5-21B-A3B-PT-UD-Q4_K_XL-layers-199/package"
Links
- Source model: unsloth/ERNIE-4.5-21B-A3B-PT-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