[!WARNING] Experimental package: artifact integrity may be validated, but runtime, split-correctness, and multimodal certification are still pending. This package is not discoverable through
meshllm/catalog@mainuntil its Hugging Face catalog PR is reviewed and merged.
GGUF layer package for running Qwen3.8-2.4T-A95B-UD-IQ2_XXS across a local Mesh LLM cluster.
This package is derived from unsloth/Qwen3.8-2.4T-A95B-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 | Q2_XXS layer package |
Model Overview
| Property | Value |
|---|---|
| Source model | unsloth/Qwen3.8-2.4T-A95B-GGUF |
| Model id | unsloth/Qwen3.8-2.4T-A95B-GGUF:UD-IQ2_XXS |
| Family | Qwen3 |
| Parameter scale | 95B |
| Quantization | Q2_XXS |
| Layer count | 93 |
| Activation width | 8192 |
| Package size | 612.4 GB |
| Source file | UD-IQ2_XXS/Qwen3.8-2.4T-A95B-UD-IQ2_XXS-00001-of-00015.gguf |
| Package repo | meshllm/Qwen3.8-2.4T-A95B-UD-IQ2_XXS-layers |
| License | other from unsloth/Qwen3.8-2.4T-A95B-GGUF |
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/Qwen3.8-2.4T-A95B-GGUF.
Quickstart
# Run this on each machine that should contribute memory/compute.
mesh-llm serve --model "meshllm/Qwen3.8-2.4T-A95B-UD-IQ2_XXS-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/Qwen3.8-2.4T-A95B-GGUF:UD-IQ2_XXS",
"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/Qwen3.8-2.4T-A95B-GGUF@567d3e6ac26c5474b18311e619c04350fb9a5556/UD-IQ2_XXS/Qwen3.8-2.4T-A95B-UD-IQ2_XXS-00001-of-00015.gguf |
| Source revision | 567d3e6ac26c5474b18311e619c04350fb9a5556 |
| Source SHA-256 | b3e4b5504945603a85b93fdca708cfdc22163c47924fd7d4c168bf60c7b88478 |
| Skippy ABI | 0.1.38 |
| Package manifest SHA-256 | efd53c02fc96c94a5ef4ddba2baaf66e2a77835536c88bdf7d018680da459103 |
What Is Included
| Artifact | Path | Contents | SHA-256 |
|---|---|---|---|
| Manifest | model-package.json | Package schema, source identity, checksums | efd53c02fc96c94a5ef4ddba2baaf66e2a77835536c88bdf7d018680da459103 |
| Metadata | shared/metadata.gguf | 0 tensors, 10.4 MB | 37ba773d94b6bc4db22179e70fa3bcde70aa779a45219f1a8c82c2c15dd5d805 |
| Embeddings | shared/embeddings.gguf | 1 tensors, 1.1 GB | 54cfaeb575af39999968da288c197f20c79d12820622441ecbd7a3bac80c1718 |
| Output head | shared/output.gguf | 2 tensors, 1.1 GB | 802ba4a41b26234c58937810a32db7c3ac4148ab8b0dfcdac468a59d8baccdf0 |
| Transformer layers | layers/layer-*.gguf | 93 layer artifacts, 1699 tensors, 610.3 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-IQ2_XXS/Qwen3.8-2.4T-A95B-UD-IQ2_XXS-00001-of-00015.gguf" --out-dir "/tmp/meshllm-layer-job-meshllm_Qwen3.8-2.4T-A95B-UD-IQ2_XXS-layers-1/package"
Links
- Source model: unsloth/Qwen3.8-2.4T-A95B-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