---
title: gemma-4-e2b-it-OptiQ-4bit-decision
canonical_url: "https://www.modelscope.cn/models/mlx-community/gemma-4-e2b-it-OptiQ-4bit-decision"
md_url: "https://www.modelscope.cn/models/mlx-community/gemma-4-e2b-it-OptiQ-4bit-decision.md"
repository: mlx-community/gemma-4-e2b-it-OptiQ-4bit-decision
last_updated: 2026-10-23
license: gemma
pipeline_tag: text-classification
tasks:
  - text-classification
model_type:
  - gemma4
architectures:
  - Gemma4ForConditionalGeneration
base_model:
  - mlx-community/gemma-4-e2b-it-OptiQ-4bit
base_model_relation: quantized
parameters: 1.1B
tensor_type:
  - U32
  - BF16
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 106
stars: 0
tags:
  - mlx
  - decision-model
  - typed-decisions
  - optiq
  - gemma-4
  - apple-silicon
---

# gemma-4-e2b-it-OptiQ-4bit-decision

> gemma-4-e2b-it-OptiQ-4bit-decision - mlx-community 在 ModelScope 开源的模型。Gemma-4 E2B decision model (OptiQ 4-bit)

mlx-community/gemma-4-e2b-it-OptiQ-4bit-decision 是 ModelScope 魔搭社区上的 1.1B 参数text-classification模型，采用 gemma 许可，基于 mlx-community/gemma-4-e2b-it-OptiQ-4bit 构建。

- **Repository**: mlx-community/gemma-4-e2b-it-OptiQ-4bit-decision
- **License**: gemma
- **Tasks**: text-classification
- **Parameters**: 1.1B
- **Base model**: mlx-community/gemma-4-e2b-it-OptiQ-4bit
- **Tags**: mlx, decision-model, typed-decisions, optiq, gemma-4, apple-silicon
- **Downloads**: 106
- **Stars**: 0
- **Last updated**: 2026-10-23

Source: https://www.modelscope.cn/models/mlx-community/gemma-4-e2b-it-OptiQ-4bit-decision

---

# Gemma-4 E2B decision model (OptiQ 4-bit)

[gemma-4-e2b-it-OptiQ-4bit](https://huggingface.co/mlx-community/gemma-4-e2b-it-OptiQ-4bit) turned into a decision model with `optiq lora train --decision`: a LoRA on the backbone and a new joint schema head that scores noul, choice and score questions in one pass. It runs on Apple Silicon through MLX and OptiQ (`optiq serve` exposes `/v1/systemone`).

## Typed Decisions result

Scored on the full 400-case test split (2,000 decisions) with the benchmark's own metrics.

| Model | Accuracy | KL from gold ↓ | Brier ↓ | ECE ↓ | p50 latency |
|---|---|---|---|---|---|
| **This model** | **0.802** | **0.072** | **0.039** | 0.181 | 0.99 s |
| [Qwen3.5-0.8B-decision](https://huggingface.co/mlx-community/Qwen3.5-0.8B-decision) | 0.772 | 0.097 | 0.054 | 0.177 | 0.34 s |
| [clef-OptiQ-4bit](https://huggingface.co/mlx-community/clef-OptiQ-4bit) (27B) | 0.719 | 0.189 | 0.102 | 0.118 | 9.7 s |
| Always answer the train-split average | 0.489 | 0.327 | 0.181 | – | – |

Latency is one request per case with all five questions, on an Apple M3 Max.

This model was fitted on the benchmark's own train split and nothing else: 1,080 records, 3 epochs, LoRA rank 16. The test cases are disjoint from it, but they come from the same four workflows, so expect less on decisions of a different kind. The Clef models were trained on other data. Top-label calibration (ECE) is no better than Clef's.

## Use

```bash
pip install -U mlx-optiq
optiq serve --model mlx-community/gemma-4-e2b-it-OptiQ-4bit-decision
```

Size on disk is 5.4 GB (4-bit backbone, vision sidecar, LoRA adapter and head).

Trained and scored with [OptiQ](https://mlx-optiq.com).