HF model

dahara1 / Qwen3.5-4B-UD-japanese-imatrix

Qwen3.5-4B‑UD‑japanese‑imatrix by dahara1 is a top‑tier, Japanese‑focused GGUF model using Unsloth Dynamic Quantization 2.0 with extensive Japanese calibration and community‑fixed bugs, runnable on llama.cpp with at least 8 GB RAM and 3 GB disk even without GPU.

Description in other languages

Statistics

Statistics as of Nov 16, 2026

Hugging Face downloads (last 30 days)
5,000
Hugging Face likes
6
Listed
Mar 22, 2026

Downloads (30-day window) over time

05k10kNovSepJulMayMarJan 20265,000
Show as a table
MonthDL (30d)
Dec 2025—
Jan 2026—
Feb 2026—
Mar 202610,000
Apr 20267,000
May 20264,000
Jun 2026—
Jul 2026—
Aug 20261,000
Sep 20262,000
Oct 2026—
Nov 20265,000

Downloads over the 30 days before the last snapshot of each month.

Popular in the same category