ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF — new model trending #30 on Hugging Face

| | SWE-bench Verified | LiveCodeBench v6 |

|---|---|---|

| BF16 base (354 GB) | 82.80 | 87.43 |

| **GSQ-RCO Coder (58.4 GB)** | **75.60** | **86.28** |

| retained | 91.3% | 98.7% |

All numbers in this card were measured at **xhigh reasoning effort**.

This is an experimental release and feedback is welcome, particularly on capabilities that were not represented in the calibration mixture.

LiveCodeBench v6 is largely preserved, at 98.7% of the base score. SWE-bench Verified, which evaluates extended agentic trajectories over real repositories, retains 91.3%. The gap between the two is consistent with the expectation that sustained multi-step tasks accumulate error across turns, and therefore tolerate capacity reduction less well than single-problem code generation.

---

## What Expert Pruning Is

A mixture-of-experts layer holds many independent feed-forward blocks and a small router that picks a few per token. Flash-Next has **512 experts per layer across 48 layers, with 10 active per token**, so the experts hold the overwhelming majority of the parameters while only a fraction runs on any given token.

Expert pruning removes entire experts from the model. It is therefore distinct from quantization, which retains every parameter and reduces the precision at which it is stored. Pruning removes parameters from the model outright, reducing both file size and resident memory, and restricts the router to a smaller candidate pool. The two techniques are complementary and are combined here: 50% of experts are removed, and the remaining weights are quantized to 3.5 bpw.

The central question is which experts to remove. Since experts specialise, importance is not an intrinsic property of an expert but is defined relative to a distribution of inputs. That distribution is supplied by the calibration data, which consequently determines which capabilities survive compression.

---

## How the Experts Are Chosen

Selection was done with **RCO** ([paper](https://arxiv.org/abs/2605.00649), [code](https://github.com/IST-DASLab/RCO)), optimizing the KL divergence between the pruned and unpruned model on calibration data, rather than ranking experts by a heuristic importance score.

The GGUF format constrains the problem in a particular way. It stores **a single expert count for the entire model**, so every layer must retain the same number of experts. A single global budget would permit the search to remove more experts from insensitive layers and fewer from sensitive ones, which generally yields a better model, but produces per-layer counts that the format cannot represent without padding that forfeits the memory saved.

RCO enforces budgets exactly by treating the feasible set as a smooth manifold, and admits **multiple simultaneous budgets**. Replacing the global budget with one budget per layer expresses the format's structural requirement directly as constraints. These constraints are mutually independent, as no expert belongs to more than one layer, so the projection and retraction steps decompose into per-layer operations and remain inexpensive.

The constraints decompose, but **the objective does not**, the loss is the KL divergence of the full model, so the gradient for any layer reflects the state of all others; the layers are optimised jointly and only budget enforcement proceeds per layer. This distinguishes the method from the common alternative of retaining a fixed number of top-ranked experts per layer under a heuristic importance score, with no joint search. A penalty-based formulation would additionally require one coefficient per layer, and these coefficients interact; RCO requires none.

---

## Vision Capability

An initial search used code and agentic calibration data only. It removed the experts on which the vision pathway depends, and image capability degraded substantially while coding scores were unaffected. This is the expected behaviour rather than a failure of the method: absent calibration data that exercises vision, the search correctly determines that those experts do not contribute to the measured objective.

The released model was searched with vision represented in the calibration mixture and retains the capability. As an informal qualitative check, the "pelican riding a bicycle" prompt is used, which requires object knowledge, spatial composition and correct SVG syntax simultaneously:

<div align="center">

<img src="assets/pelican-on-bike.svg" alt="SVG of a pelican riding a bicycle, drawn by the pruned model" width="440"/>

</div>

---

## Available Files

Two shards, both required. `llama.cpp` loads the pair when given the first.

| File | Size | Must be resident | Notes |

|---|---|---|---|

| `...-00001-of-00002.gguf` | 29.6 GB | yes | Transformer weights, 256 of 512 experts per layer, 1.89 bpw over the original parameters |

| `...-00002-of-00002.gguf` | 28.8 GB | no | Per-layer n-gram table (IQ4_NL, 4.5 bpw), neither pruned nor searched |

| **Total** | **58.4 GB** | **29.6 GB** | against 354 GB and 354 GB resident for BF16 |

The n-gram table is a lookup rather than a matmul weight, so it is held at a fixed 4.5 bpw and excluded from both the quantization search and the pruning.

### llama.cpp

```bash

hf download <this-repo> --local-dir .

llama-cli -m IQ1_M/Qwen3.8-Flash-Next-GSQ-RCO-IQ1_M-00001-of-00002.gguf \

-p "Refactor this function to be iterative." -ngl 99

```

### Strata Engine

Many users report strong results with [Strata](https://github.com/Niko1221/Strata), a third-party engine that runs these GGUFs on a single NVIDIA GPU from 8 GB of VRAM, with a one-click installer and an OpenAI-compatible endpoint on localhost.

---

## Method

| | |

|---|---|

| **GSQ** ([paper](https://arxiv.org/abs/2604.18556), [code](https://github.com/IST-DASLab/GSQ)) | Post-training scalar quantization that learns grid assignments and group scales jointly, closing most of the gap to vector quantization at low bit-widths while staying deployable in standard GGUF types. |

| **RCO** ([paper](https://arxiv.org/abs/2605.00649), [code](https://github.com/IST-DASLab/RCO)) | Budget-constrained optimization on a Riemannian manifold. Used here to choose which experts to remove under one exact budget per layer, and in the base releases to assign per-tensor quantization types. |

Both developed at the [Deep Algorithms and Systems Lab (DASLab)](https://github.com/IST-DASLab), Institute of Science and Technology Austria.

---

## Citation

```bibtex

@article{gsq2026,

title = {GSQ: Highly-Accurate Low-Precision Scalar Quantization for LLMs via Gumbel-Softmax Sampling},

author = {Dadgarnia, Alireza and Tabesh, Soroush and Nikdan, Mahdi and Helcig, Michael and Kurtic, Eldar and Kleinegger, Maximilian and Alistarh, Dan},

journal= {arXiv preprint arXiv:2604.18556},

year = {2026}

}

@article{rco2026,

title = {Model Compression with Exact Budget Constraints via Riemannian Manifolds},

author = {Helcig, Michael and Alistarh, Dan},

journal= {arXiv preprint arXiv:2605.00649},

year = {2026}

}

```

---

## Acknowledgements

We thank [Verda](https://verda.com/) and Scientific Computing at the Institute of Science and Technology Austria for the compute used to produce these models.

## License

These weights inherit the license of the base model (**Qwen3.8-Flash-Next**). The GSQ-RCO tooling is released by the Deep Algorithms and Systems Lab under its repository license.

<div align="center">

_{Built with <b>GSQ</b> and <b>RCO</b> at the <a href="https://github.com/IST-DASLab">Deep Algorithms and Systems Lab</a> &middot; Institute of Science and Technology Austria}

</div>

Text extracted automatically; images, tables and formatting may be missing. Original: https://huggingface.co/ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF