---
license: apache-2.0
base_model: z-lab/Qwen3.5-9B-DFlash
tags:
- gguf
- llama.cpp
- speculative-decoding
- dflash
- draft-model
---

# Qwopus3.5-9B-Coder-DFlash-GGUF

DFlash draft model for [Jackrong/Qwopus3.5-9B-Coder](https://huggingface.co/Jackrong/Qwopus3.5-9B-Coder), for use with upstream llama.cpp speculative decoding (`--spec-type draft-dflash`, merged in [#22105](https://github.com/ggml-org/llama.cpp/pull/22105)).

This is [z-lab/Qwen3.5-9B-DFlash](https://huggingface.co/z-lab/Qwen3.5-9B-DFlash) converted with `--target-model-dir` pointing at the Qwopus tokenizer. Qwopus extends the Qwen3.5 tokenizer with 7 added tokens (ids 248070 to 248076), so drafts converted against the base Qwen3.5-9B tokenizer do not pass the vocab compatibility check for this target. The draft GGUF carries no token embeddings or lm_head; llama.cpp shares the target model's at runtime, so the draft matches whatever Qwopus quant you serve.

Conversion and launch scripts: https://github.com/Gaurav-Gosain/qwopus-dflash

## Usage

```sh
llama-server \
  -m Qwopus3.5-9B-coder-Exp-Q3_K_M.gguf \
  -md Qwopus3.5-9B-Coder-DFlash-Q4_K_M.gguf \
  --spec-type draft-dflash --spec-draft-n-max 15 \
  -fa on --jinja -ctxcp 2
```

Qwen3.5 is a hybrid linear-attention architecture; keep `-ctxcp` low because each context checkpoint stores the full recurrent state (about 100 MB).

## Files

| file | size | note |
| --- | --- | --- |
| Qwopus3.5-9B-Coder-DFlash-Q4_K_M.gguf | 766 MB | recommended |
| Qwopus3.5-9B-Coder-DFlash-Q8_0.gguf | 1.4 GB | measured identical speed to Q4_K_M |
| Qwopus3.5-9B-Coder-DFlash-bf16.gguf | 2.6 GB | for requantizing |

## Measured (RTX 3070 8 GB, target Q3_K_M, 600-token code generation, temp 0)

| config | speed | acceptance |
| --- | --- | --- |
| no speculation | 38 tok/s | - |
| DFlash, n-max 7 | 81 tok/s | 0.56 |
| DFlash, n-max 15 | 127 tok/s | 0.34 |

Acceptance drops to about 0.15 on freeform prose, still roughly 1.8x. Both models plus buffers need about 6.5 GB free VRAM; if the target spills layers to CPU, speculation becomes a net loss.