serverAppConfigs/docs/DFlash DSpark Speculative Decoding.md

# DFlash and DSpark Speculative Decoding

## Overview

DFlash and DSpark are the two newest speculative decoding methods. They draft many tokens in one parallel pass instead of one token at a time. This reduces decode latency on memory-bound workloads.

DFlash comes from Z Lab at UC Berkeley. DSpark comes from DeepSeek. Both work with vLLM and llama.cpp. Both have draft models for the Qwen3.6 models that Dankserver and Dankclaw run.

## How DFlash Works

DFlash is a lightweight block diffusion model. It predicts a block of tokens in a single forward pass. Key design points:

- The draft model reuses the embedding and LM head from the target model.
- Only the intermediate layers are trained.
- It fuses target-model hidden features into the KV projections of every draft layer.
- A 16-token DFlash drafter has lower latency than a 1-layer EAGLE-3 drafter producing 8 tokens.

Results from the DFlash paper (arXiv:2602.06036):

- Over 6x lossless speedup on Qwen3-8B.
- Up to 2.5x higher speedup than EAGLE-3.
- Diffusion drafting cost is flat, so block size does not increase cost.

## How DSpark Works

DSpark is a semi-autoregressive extension of DFlash from DeepSeek (arXiv:2607.05147). It keeps the DFlash parallel diffusion backbone and adds a lightweight sequential Markov head. This models inter-token dependencies inside each block. It also uses confidence-scheduled verification, which stops verifying tokens that have a high rejection risk.

Results from the DSpark paper:

- 60-85% faster per-user generation than MTP-1 on DeepSeek-V4 production traffic.
- Useful when parallel drafts decay in acceptance rate.

Both DFlash and DSpark are lossless under rejection sampling. Use the strict rejection sampler.

## Runtime Support

| Runtime | DFlash | DSpark | MTP |
|---------|--------|--------|-----|
| vLLM | yes (v0.20.1+) | yes | yes |
| llama.cpp | yes | yes | yes |
| SGLang | yes | no | yes |

### vLLM

vLLM v0.20.1+ has core DFlash support. Use the `--speculative-config` JSON object.

DFlash on vLLM:

```
vllm serve Qwen/Qwen3.6-35B-A3B \
    --speculative-config '{"method": "dflash", "model": "z-lab/Qwen3.6-35B-A3B-DFlash", "num_speculative_tokens": 15}' \
    --attention-backend flash_attn \
    --max-num-batched-tokens 32768
```

Legacy flags also work:

```
vllm serve Qwen/Qwen3.6-35B-A3B \
    --spec-method dflash \
    --spec-model z-lab/Qwen3.6-35B-A3B-DFlash \
    --spec-tokens 8
```

DFlash requires a non-causal attention backend. Set `"attention_backend": "flash_attn"` in the speculative config if needed. On ROCm, check that the flash_attn backend is available before using DFlash.

DSpark on vLLM:

```
vllm serve <target-model> \
    --speculative-config '{"method": "dspark", "model": "<dspark-draft>", "num_speculative_tokens": 15}'
```

### llama.cpp

llama.cpp supports DFlash, DSpark, and MTP through `--spec-type`. When you pass `--spec-draft-hf`, llama.cpp auto-discovers the sidecar files and infers the type. The priority order is DSpark, DFlash, EAGLE-3, MTP.

`--spec-type` accepts a **comma-separated list** of speculative implementations. `common_speculative_init` (in `common/speculative.cpp`) builds a vector of implementations and chains drafting through all of them (`common_speculative_draft_params.drafting` is "used to chain the drafts through all the available implementations"). This means you can **layer ngram-mod on top of a DFlash draft model** in one server process:

```
llama-server --hf-repo <target-GGUF> \
    --spec-draft-hf <dflash-draft-GGUF> \
    --spec-type draft-dflash,ngram-mod
```

The DFlash drafter proposes, then ngram-mod runs as an additional implementation in the chain. Verified against upstream `common/speculative.cpp` (b10133+).

DFlash on llama.cpp:

```
llama-server --hf-repo <target-GGUF> \
    --spec-draft-hf <draft-GGUF> \
    --spec-type draft-dflash \
    --spec-draft-n-max 16
```

DSpark on llama.cpp:

```
llama-server --hf-repo <target-GGUF> \
    --spec-draft-hf <draft-GGUF> \
    --spec-type draft-dspark
```

MTP on llama.cpp:

```
llama-server --hf-repo <target-with-MTP> \
    --spec-type draft-mtp \
    --spec-draft-n-max 2
```

Note: the flag was renamed from `--spec-type mtp` to `--spec-type draft-mtp` in May 2026 (PR #22673, build b9180+). An old flag silently disables speculative decoding.

### MTP drafter availability for Qwen3.6-27B

MTP is not a separate drafter model. It is a "nextn" head that must live in the target GGUF (or be loaded as a draft alongside it). llama.cpp's `draft-mtp` implementation is fully wired for the `qwen35` and `qwen35moe` architectures (PR #22673).

Local check on Dankclaw (2026-08-13), tensor listing via GGUF header parse:

| File | Architecture | Tensors | MTP tensors |
|------|--------------|---------|-------------|
| `Qwen3.6-27B-Uncensored-HauhauCS-Aggressive-Q8_K_P.gguf` | `qwen35` | 851 | 0 |
| `Qwen3.6-27B-DFlash-Q8_0.gguf` (draft) | `dflash` | 58 | 0 |

The HauhauCS uncensored GGUFs ship without the MTP head. To use MTP with them, load an MTP-only draft GGUF from the base Qwen3.6-27B model as a sidecar:

```
llama-server --model <HauhauCS-Q8_K_P.gguf> \
    --spec-draft-model <MTP-only-draft.gguf> \
    --spec-type draft-mtp \
    --spec-draft-n-max 2
```

Compatible MTP-only draft GGUFs on Hugging Face:

| Repo | Notes |
|------|-------|
| `a4lg/Qwen3.6-27B-MTP-ONLY-GGUF` | MTP-only subset of the base model; pairs with derivative/abliterated targets via `--spec-draft-model` |
| `lym00/Qwen3.6-27B-MTP-ONLY-GGUF` | Same purpose; built from `convert_hf_to_gguf.py --mtp` |
| `ggml-org/Qwen3.6-27B-MTP-GGUF` | Full model with MTP head (replaces the target, not a sidecar) |
| `hyperhuzaifa/Qwen3.6-27B-Uncensored-MTP-GGUF` | HauhauCS Q4_K_P with the MTP head grafted in (replaces the target) |

Speculative decoding is lossless under rejection sampling: the target verifies every draft token. A draft head from the base model works with a derivative target; wrong proposals just lower the acceptance rate. The accepted MTP draft lengths are 2-3 tokens; the "MTP Defaults Are a Trap" writeup (thefrontierlab.ai) documents the tuning pitfalls.

On Dankclaw, `draft-dflash` remains the benchmarked winner (20.2 t/s at 4k vs 16.9 for DFlash+ngram and 8.7 for ngram-only). MTP is a viable alternative to test, not a bug fix: the parallel-stream releases investigated on 2026-08-13 were client-side aborts and a Dankserver crash, not speculative decoding behavior (see [[deployment]]).

## Draft Models

DFlash draft model sizes for the models on these servers:

| Target model | DFlash draft | Draft size |
|--------------|--------------|------------|
| Qwen3.6-35B-A3B | z-lab/Qwen3.6-35B-A3B-DFlash | 0.4B |
| Qwen3.6-27B | z-lab/Qwen3.6-27B-DFlash | 2B |
| Qwen3.5-35B-A3B | z-lab/Qwen3.5-35B-A3B-DFlash | 0.4B |
| Qwen3.5-27B | z-lab/Qwen3.5-27B-DFlash | 2B |

The DFlash draft is not a standalone model. It must pair with its target model. The model cards recommend block size 8 for higher concurrency and 16 for single-stream throughput.

## Deployment on Dankserver (NVIDIA RTX 3090)

## Deployment on Dankserver (NVIDIA RTX 3090)

The active Dankserver deployment runs **llama.cpp** (`frank-default-small`), not vLLM. It serves the HauhauCS Qwen3.6-35B-A3B uncensored `Q6_K_P` on port 7770 using **ngram-mod** speculative decoding (n_min=48 / n_max=64 / n_match=24) with 3 slots at 200000 context each and MoE experts on CPU. See [[deployment]] and `scripts/serve-llama-frank-small.sh`.

> **DFlash retired from active service.** `DFlash` (draft-dflash) was tested initially but did not deliver the desired speed gains under the `frank-default-small` workload (see ngram-mod benchmarks below). The `Alittlehammmer/Qwen3.6-35B-A3B-DFlash` draft GGUF remains in `/home/dank/models/` as reference material only.

The vLLM material below documents the archived AWQ/DFlash path on the same hardware. The RTX 3090 is Ampere (compute capability 8.6). Use these conditions:

- FP8 KV cache is valid on CUDA 12.
- FP8 weights compute through a fallback path; no native FP8 tensor cores.
- The DFlash draft model (0.4B) fits easily in 24 GB VRAM alongside the target.

Working DFlash flags for this server:

```
--spec-method dflash
--spec-model z-lab/Qwen3.6-35B-A3B-DFlash
--spec-tokens 8
--kv-cache-dtype fp8
```

Example DFlash + AWQ on the RTX 3090:

```
docker run --rm --name vllm-qwen36-dflash \
    --gpus all --ipc=host \
    -p 8000:8000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    docker.io/vllm/vllm-openai:latest \
    cyankiwi/Qwen3.6-35B-A3B-AWQ-4bit \
    --speculative-config '{"method": "dflash", "model": "z-lab/Qwen3.6-35B-A3B-DFlash", "num_speculative_tokens": 8}' \
    --attention-backend flash_attn \
    --gpu-memory-utilization 0.95
```

## Deployment on Dankclaw (AMD Strix Halo gfx1151)

Dankclaw now runs llama.cpp as its runtime for `frank-default-large` (port 7770), serving the HauhauCS Qwen3.6-27B uncensored `Q8_K_P` model with an **MTP sidecar draft** on the ROCm (HIP) backend. DFlash was retired on 2026-08-14.

Current facts:

- Main model: HauhauCS/Qwen3.6-27B-Uncensored-HauhauCS-Aggressive `Q8_K_P` (dense 27B).
- Draft model: a4lg/Qwen3.6-27B-MTP-ONLY `Q8_0` (MTP head from the base model, loaded as a sidecar).
- Spec command: `--spec-type draft-mtp` with `--spec-draft-n-max 3`.
- Backend `--device ROCm0`; `--gpu-layers all --fit off`; KV `q8_0` K / `q8_0` V.
- Context: 3 x 200700 = 602100; batch 8192 / ubatch 8192.

Why DFlash was retired: DFlash causes premature EOS on a generating slot when another task lands with high prompt-cache reuse (LCP ~0.99). See [[frank-default-large-benchmark]] for the incident writeup and the MTP-vs-ngram A/B. The DFlash draft GGUF (`Qwen3.6-27B-DFlash-Q8_0.gguf`) remains in `/home/dank/models/` as reference only.

## Notes

- Speculative decoding helps low-QPS, memory-bound, latency-focused workloads. It helps less at high QPS.
- DFlash needs a non-causal attention backend; the ROCm flash_attn backend must be validated before production use.
- The old `--swap-space` flag no longer exists in current vLLM. Use `--kv-offloading-size`.
- `--speculative-config` is the preferred CLI. Legacy `--spec-method`, `--spec-model`, and `--spec-tokens` flags still exist.
- Use `rejection_sample_method` default `strict` for lossless output.

## See Also

- [[Quantization by Hardware Architecture]]
- [[Running Qwen3.6 35B A3B FP8 on Dankserver]]
- [[Running Qwen3.6 27B FP8 on Dankclaw]]
- [[Dankclaw System Specs]]
- [[Dankserver System Specs]]
- [[VLLM Overview]]
- [[deployment]]