You can deploy this model and start earning money today!

CMBTech/
CMB-Qwen3-30B-A3B-DSpark-preview

Name: CMB-Qwen3-30B-A3B-DSpark-preview
Availability: InStock

safetensors

custom_code

apache-2.0

Views

No views yet

CMB-Qwen3-30B-A3B-DSpark-preview

Model Overview

Target Model: Qwen/Qwen3-30B-A3B
Speculative Decoding Algorithm: DSpark
Release Date: 08/17/2026
Version: preview
Model Developer: Architecture Management Team, Information Technology Department, China Merchants Bank (CMB)

Model Introduction

This model is implemented based on the DSpark speculative decoding algorithm, with Qwen/Qwen3-30B-A3B as the target model.

This model was trained and open-sourced by the Architecture Management Team of the Information Technology Department at China Merchants Bank. Built upon the Qwen3-30B-A3B architecture, it was trained using the CMB-Eagle-Mix-330k dataset regenerated in `

mode (withblock_size=8`).

This model should be used with the chat template of Qwen/Qwen3-30B-A3B and invoked via the /chat/completions API endpoint.

Deployment & Usage

Using vLLM

Note: When using vLLM, you must enable the dedicated configuration file config_vllm.json for this model. Before running the command below, please rename config_vllm.json in the model directory to config.json for it to take effect.

bash

vllm serve Qwen/Qwen3-30B-A3B \
  --speculative-config '{
    "method": "dspark",
    "num_speculative_tokens": 7,
    "model": "<our-dspark-model-path>"
  }'

Evaluations

Evaluation Results (num_speculative_tokens = 7)

Category	Source Dataset	Sample Size	Avg. Accept Length
Math Reasoning	GSM8K	200	4.6
Math Reasoning	AIME2025_Prompt	30	4.14
Code Generation	HumanEval	200	4.04
Multi-turn Dialogue	MTBench_Prompt	80	3.33
General Comprehensive	Magpie-Qwen2-Pro-200K	200	2.72

Detailed Experimental Configuration

Temperature: [0]
Top_p: [0.95]
Top_k: [20]
Repetitions: [1.0]
Hardware: [2×H800]

CMBTech/CMB-Qwen3-30B-A3B-DSpark-preview