---
title: Qwen3.8-27B-DFlash2-NPU2
canonical_url: "https://www.modelscope.ai/models/AMD/Qwen3.8-27B-DFlash2-NPU2"
md_url: "https://www.modelscope.ai/models/AMD/Qwen3.8-27B-DFlash2-NPU2.md"
repository: AMD/Qwen3.8-27B-DFlash2-NPU2
last_updated: 2026-11-05
license: apache-2.0
model_type:
  - qwen3
architectures:
  - DFlash2DraftModel
base_model:
  - z-lab/Qwen3.8-27B-DFlash2
  - Qwen/Qwen3.8-27B
base_model_relation: finetune
library_name:
  - other
frameworks:
  - other
downloads: 13
stars: 0
tags:
  - speculative-decoding
  - dflash2
  - fastflowlm
  - npu
---

# Qwen3.8-27B-DFlash2-NPU2

> Qwen3.8-27B-DFlash2-NPU2 - An open-source model by AMD on ModelScope. FastFlowLM build: Qwen3.8-27B-DFlash2-NPU2

AMD/Qwen3.8-27B-DFlash2-NPU2 is a machine learning model on ModelScope. licensed under apache-2.0. derived from z-lab/Qwen3.8-27B-DFlash2、Qwen/Qwen3.8-27B.

- **Repository**: AMD/Qwen3.8-27B-DFlash2-NPU2
- **License**: apache-2.0
- **Base model**: z-lab/Qwen3.8-27B-DFlash2, Qwen/Qwen3.8-27B
- **Tags**: speculative-decoding, dflash2, fastflowlm, npu
- **Downloads**: 13
- **Stars**: 0
- **Last updated**: 2026-11-05

Source: https://www.modelscope.ai/models/AMD/Qwen3.8-27B-DFlash2-NPU2

---

# FastFlowLM build: Qwen3.8-27B-DFlash2-NPU2

The **DFlash 2 speculative-decoding draft** of Qwen3.8-27B ([z-lab/Qwen3.8-27B-DFlash2](https://huggingface.co/z-lab/Qwen3.8-27B-DFlash2),
a mirror of [incoai/Qwen3.8-27B-DFlash2](https://huggingface.co/incoai/Qwen3.8-27B-DFlash2)), converted for the AMD Ryzen AI NPU and run by
[FastFlowLM](https://github.com/FastFlowLM/FastFlowLM) (`flm`). It does nothing on its own: it is the second half of the
`qwen3.8-dflash2:27b` tag, whose main weights are the existing **`qwen3.8-mtp:27b`** package (`Qwen3.8-27B-NPU2`, unchanged).
`flm pull qwen3.8-dflash2:27b` downloads both.

```shell
flm run qwen3.8-dflash2:27b
```

Needs FastFlowLM 1.0.12 or later. `FLM_DFLASH=0` loads the model's own MTP head instead (the tag then decodes like `qwen3.8-mtp:27b`).

| | |
|---|---|
| Upstream | [z-lab/Qwen3.8-27B-DFlash2](https://huggingface.co/z-lab/Qwen3.8-27B-DFlash2) (BF16 safetensors, the source of the weights here), pinned at revision `50307d4c` |
| Requires | `qwen3.8-mtp:27b` (main model, tokenizer, chat template, vision tower, kernels) |
| License | Apache-2.0 (`LICENSE`); the main model's license applies to the main weights |
| Draft | 5 Qwen3-style layers, hidden 5120, FFN 17408, 32 query / 8 key-value heads of 128, sliding window 2048, block of 8 tokens, two-tap dynamic convolutions, candidate selector (top 16, rank 256); 2 B parameters; shares the target's embedding table and output head |
| Weights | `dflash_weight.q4nx`, 1.39 GB: `fc`, q/k/v/o and MLP matrices in Q4_K (about 0.07 relative L2 per matrix); norms, the dynamic-convolution tensors, the selector projection and the two 248320 x 256 selector codebooks in BF16 |
| Taps | features of target layers 5, 19, 33, 47 and 61 (zero-based layer outputs), fused by the draft's `fc` layer |

### How it is used

Each step the draft proposes 7 tokens in one pass and picks one coherent path through its top 16 candidates per position; the main model checks
all 7 in one pass and keeps the longest prefix it agrees with, plus one token of its own. The draft runs on the same NPU kernels as the main
model (no extra bitstream): its matrices on the fused Q4_K dequantize + matmul engine, its output projection on the 8-row language-model head,
its attention, convolutions and selector on the CPU.

### Measured draft quality

Teacher-forced replay of one run of the real model (NPU, greedy, 300 tokens, thinking off): for every step the draft is asked for 7 tokens and
compared with what the main model actually emitted next. Mean number of the 7 drafted tokens kept per step (the model then adds one):

| Draft | code prompt (52 steps) | prose prompt (66 steps) |
|---|--:|--:|
| BF16 weights, fp32 math (z-lab's reference code) | 5.02 | 3.36 |
| this package's Q4_K weights, fp32 math | 4.87 | 3.49 |
| this package on the NPU (bf16 activations) | 4.60 | 3.41 |

Quantizing the draft costs nothing measurable beyond the noise of 50 to 70 steps; the bf16 activations of the NPU cost a few percent on code.
The published DFlash 2 acceptance lengths (which count the added token) are 4.1 on MT-Bench and 4.4 to 5.5 on code and math; this engine keeps
about 3.1 on open prose and about 6.6 on a short code prompt.

### FastFlowLM measurements

Ryzen AI MAX+ box, Linux, default power mode, `flm serve`, 160 generated tokens, greedy, ten short prompts, decode tokens/s:

| | `qwen3.8-mtp:27b` | `qwen3.8-dflash2:27b` |
|---|--:|--:|
| mean | 3.23 | 3.74 (1.16x) |
| per prompt | | 0.87x to 1.47x |

The speedup depends on how predictable the text is (code and structured answers gain the most). One step takes about 1.2 s with either draft
(0.9 s to check the 8 positions, about 0.27 s to draft). The NPU checks the 8 positions in one pass, which rounds differently from decoding one
token at a time, so greedy output can differ from `qwen3.8-mtp:27b` at near-ties (4 of 5 short greedy answers were identical to it, 5 of 10
long ones). Drafted tokens are compared with the model's own most likely token and skip the sampler, as with the MTP head.

### Credits

Draft model: Inco AI, DFlash 2 ([blog](https://inco.ai/blog/dflash2/)); original DFlash: Chen, Liang, Liu (ICML 2026),
[z-lab/dflash](https://github.com/z-lab/dflash).