Sponsored by Vmake Labs Video Enhancer
Toolify Toolify logo
AI Models
Qwen3.8-27B-DFlash2-NVFP4-modelopt

hamichok / Qwen3.8-27B-DFlash2-NVFP4-modelopt

Run on huggingface.co
huggingface.co
Total runs: 393
24-hour runs: 2
7-day runs: -110
30-day runs: 42
Model's Last Updated: August 30 2026

Introduction of Qwen3.8-27B-DFlash2-NVFP4-modelopt

Model Details of Qwen3.8-27B-DFlash2-NVFP4-modelopt

Qwen3.8-27B-DFlash2-NVFP4-modelopt

Re-quantized DFlash2 draft for sglang's DFLASH speculative decoding, used by hamichok/Qwen3.8-27B-NVFP4-RTX5090-LMHead4 .

This is a re-quantization of incoai/Qwen3.8-27B-DFlash2 (BF16) to modelopt-NVFP4 so sglang can load it as a speculative draft. The upstream compressed-tensors NVFP4 variant does not load in sglang; this one does.

Why this exists
  • DFLASH needs a draft the target lm_head -quantized checkpoint can verify; the quantized head breaks DSPARK but not DFLASH.
  • fc and q/k/v stay dense ; only o_proj + MLP projections are quantized (the dflash.py loader shape-checks fc.weight as dense).
Usage (serve command)

Point --speculative-draft-model-path at this repo and --speculative-draft-model-quantization modelopt_fp4 , with --speculative-dflash-block-size 6 (measured throughput peak). See the main model card for the full recipe and measurements.

Provenance

Built with quantize-dflash2.py (Apache-2.0, from the main recipe repo). Base: incoai/Qwen3.8-27B-DFlash2 (Apache-2.0).

Runs of hamichok Qwen3.8-27B-DFlash2-NVFP4-modelopt on huggingface.co

393
Total runs
2
24-hour runs
7
3-day runs
-110
7-day runs
42
30-day runs

More Information About Qwen3.8-27B-DFlash2-NVFP4-modelopt huggingface.co Model

More Qwen3.8-27B-DFlash2-NVFP4-modelopt license Visit here:

https://choosealicense.com/licenses/apache-2.0

Qwen3.8-27B-DFlash2-NVFP4-modelopt huggingface.co

Qwen3.8-27B-DFlash2-NVFP4-modelopt huggingface.co is an AI model on huggingface.co that provides Qwen3.8-27B-DFlash2-NVFP4-modelopt's model effect (), which can be used instantly with this hamichok Qwen3.8-27B-DFlash2-NVFP4-modelopt model. huggingface.co supports a free trial of the Qwen3.8-27B-DFlash2-NVFP4-modelopt model, and also provides paid use of the Qwen3.8-27B-DFlash2-NVFP4-modelopt. Support call Qwen3.8-27B-DFlash2-NVFP4-modelopt model through api, including Node.js, Python, http.

Qwen3.8-27B-DFlash2-NVFP4-modelopt huggingface.co Url

https://huggingface.co/hamichok/Qwen3.8-27B-DFlash2-NVFP4-modelopt

hamichok Qwen3.8-27B-DFlash2-NVFP4-modelopt online free

Qwen3.8-27B-DFlash2-NVFP4-modelopt huggingface.co is an online trial and call api platform, which integrates Qwen3.8-27B-DFlash2-NVFP4-modelopt's modeling effects, including api services, and provides a free online trial of Qwen3.8-27B-DFlash2-NVFP4-modelopt, you can try Qwen3.8-27B-DFlash2-NVFP4-modelopt online for free by clicking the link below.

hamichok Qwen3.8-27B-DFlash2-NVFP4-modelopt online free url in huggingface.co:

https://huggingface.co/hamichok/Qwen3.8-27B-DFlash2-NVFP4-modelopt

Qwen3.8-27B-DFlash2-NVFP4-modelopt install

Qwen3.8-27B-DFlash2-NVFP4-modelopt is an open source model from GitHub that offers a free installation service, and any user can find Qwen3.8-27B-DFlash2-NVFP4-modelopt on GitHub to install. At the same time, huggingface.co provides the effect of Qwen3.8-27B-DFlash2-NVFP4-modelopt install, users can directly use Qwen3.8-27B-DFlash2-NVFP4-modelopt installed effect in huggingface.co for debugging and trial. It also supports api for free installation.

Qwen3.8-27B-DFlash2-NVFP4-modelopt install url in huggingface.co:

https://huggingface.co/hamichok/Qwen3.8-27B-DFlash2-NVFP4-modelopt

Url of Qwen3.8-27B-DFlash2-NVFP4-modelopt

Qwen3.8-27B-DFlash2-NVFP4-modelopt huggingface.co Url

Provider of Qwen3.8-27B-DFlash2-NVFP4-modelopt huggingface.co

hamichok
ORGANIZATIONS
huggingface.co

Other API from hamichok