inclusionAI/Ling-3.0-flash-dspark

Ling3-DSpark

A DSpark speculator for Ling3. DSpark extends DFlash with target-model auxiliary features and a confidence head that dynamically chooses the number of draft tokens. The model was trained with SpecForge and is served with SGLang.

Model specifications

• Target model: Ling-3.0-flash
• Draft parameters: 1,363,707,905 (1.36B)
• Draft weight dtype: BF16
• Hidden size: 2,560
• Transformer layers: 5 full-attention layers
• Attention: MHA with 32 query heads and 32 key/value heads
• Target auxiliary feature layers: 1, 11, 23, 29, 35
• Confidence head: vanilla Markov head, rank 256
• DSpark block size: 8 draft tokens (verify width 9, including the target bonus token)
• Maximum position embeddings: 262,144

Acceptance length

Acceptance length is the mean number of tokens accepted per speculative verification step, including the target bonus token.

| Workload | Acceptance length | |---|---:| | GSM8K | 6.40 | | MATH-500 | 6.29 | | AIME 2025 | 5.56 | | HumanEval | 6.57 | | MBPP | 6.34 | | LiveCodeBench | 5.33 | | MT-Bench | 3.92 | | Alpaca | 3.51 | | Arena-Hard-v2 | 3.72 |

The macro mean across the nine workload means is 5.29.

Serving with SGLang

Launch recipes for this draft on every supported hardware/quantization cell — including the required --linear-replayssm-cache-len sizing — with measured speed and accuracy, are in the SGLang Ling-3.0-flash cookbook.

Use an SGLang version with DSPARK support. Replace the model paths and tensor-parallel size with values appropriate for your deployment:

sglang serve \
--trust-remote-code \
--model-path \
--tp-size \
--speculative-algorithm DSPARK \
--speculative-draft-model-path \
......