Downloads · 30 days
42
58% of all-time downloads
Qwen3.8-27B-test-DFlash2-v0.2
vvsotnikov/Qwen3.8-27B-test-DFlash2-v0.2
Qwen3.8-27B-test-DFlash2-v0.2 is a machine learning model from vvsotnikov. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Use with the target above, not vanilla Qwen. This repository contains BF16 draft weights; --draft-bits 4 quantizes the drafter at load time.
All-time downloads
73
Public
Parameters
1.9B
3.8 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors3.8 GB · 100%
At a glance
- Model type
- qwen3
- Access
- Public
- Created
- Sep 22, 2026
- Updated
- Sep 22, 2026
- SHA
- 68e29209
Model card
From the Hugging Face model README
How to launch (Apple Silicon / MLX)
python3 -m venv .venv
source .venv/bin/activate
pip install "dflash[local] @ git+https://github.com/z-lab/dflash.git@07ebd93db9f472af339b644bb70221ad8428328a"
dflash generate mlx "Write a Python function that checks whether a string is a palindrome." \
--model vvsotnikov/Qwen3.8-27B-test-MLX-4bit \
--draft vvsotnikov/Qwen3.8-27B-test-DFlash2-v0.2 \
--draft-bits 4 --block-size 8 \
--reasoning low --temperature 1.0 --top-p 0.95 --top-k 20 \
--max-new-tokens 256
Use with the target above, not vanilla Qwen. This repository contains BF16 draft
weights; --draft-bits 4 quantizes the drafter at load time.
- Type
- qwen3
- Created
- Sep 22, 2026
- Updated
- Sep 22, 2026