Qwen3.8-27B-MixedInt2-ultra-lite

A mixed-precision AutoRound quantized version of Qwen/Qwen3.8-27B, optimized to significantly reduce memory requirements while preserving the quality of the original model.

Base Model

Base model: Qwen/Qwen3.8-27B

This model is a quantized version of the original Qwen3.8-27B checkpoint. It is not a fine-tune, merge, or distillation.

Quantization

Quantization was performed using Intel AutoRound with a custom mixed-precision quantization configuration.

The quantization recipe was designed to balance:

• Model quality
• Memory efficiency
• Inference performance
• vLLM compatibility

Some model components are intentionally retained at higher precision where appropriate.

Property	Value
Quantization framework	Intel AutoRound
Quantization type	Custom Mixed-Precision INT2
Group size	32, 64, 128
Base model	Qwen/Qwen3.8-27B
Language layers	64
Vision tower	Preserved at original precision

The detailed mixed-precision allocation strategy is not included in this model card.

Evaluation

Evaluation was performed using AutoRound's evaluation interface with LM Evaluation Harness.

The following results compare the original Qwen3.8-27B model against Qwen3.8-27B-MixedInt2-ultra-lite.

Benchmark	Metric	Qwen3.8-27B	Qwen3.8-27B-MixedInt2-ultra-lite	Difference	Recovery Rate
MMLU	acc	83.49%	78.30%	-5.19 pp	93.78%
GSM8K	exact_match (flexible)	72.86%	43.29%	-29.57 pp	59.42%
ARC-Challenge	acc_norm	—	—	—	—
BoolQ	acc	—	—	—	—
HellaSwag	acc_norm	—	—	—	—
PIQA	acc_norm	—	—	—	—
WinoGrande	acc	—	—	—	—
Average	—	—	—	—	—

MMLU Category Breakdown

MMLU Category	Qwen3.8-27B	Qwen3.8-27B-MixedInt2-ultra-lite	Difference	Recovery Rate
Humanities	77.39%	73.33%	-4.06 pp	94.75%
Other	86.03%	81.20%	-4.83 pp	94.39%
Social Sciences	90.74%	85.99%	-4.75 pp	94.77%
STEM	83.03%	75.36%	-7.67 pp	90.76%

GSM8K

Metric	Qwen3.8-27B	Qwen3.8-27B-MixedInt2-ultra-lite	Difference	Recovery Rate
Flexible Exact Match	72.86%	43.29%	-29.57 pp	59.42%
Strict Exact Match	70.36%	35.33%	-35.03 pp	50.22%

Recovery Rate represents benchmark performance relative to the original Qwen3.8-27B checkpoint. A recovery rate above 100% indicates that the quantized model scored higher than the original model in that particular evaluation. Benchmark preservation does not imply identical behavior for every prompt, multimodal workload, long-context workload, or generation setting.

Usage

This checkpoint is intended for inference engines with AutoRound quantization support, including compatible versions of vLLM.

Notes

• This is a quantized derivative of Qwen3.8-27B.
• The model uses a custom mixed-precision INT2 quantization recipe.
• The vision components are preserved at their original precision.
• This ultra-lite variant is designed to further reduce memory requirements compared with higher-bit quantized variants.
• Quantization may introduce small behavioral differences compared with the original checkpoint.

Acknowledgements

• Base model: Qwen Team
• Quantization framework: Intel AutoRound

Please refer to the original Qwen3.8-27B model card for licensing, intended usage, limitations, and other information applicable to the base model.