HF modelNEW

third-intelligence / Haas-ja-v1.0

Haas‑ja‑v1.0 is a fast, accurate Japanese speech‑recognition model using NVIDIA’s Parakeet‑TDT architecture with a FastConformer encoder and a Kimi Delta Attention prediction network, achieving a strict 6.12 % CER while outperforming commercial APIs at a fraction of their cost.

Description in other languages

Statistics

Statistics as of Nov 16, 2026

Hugging Face downloads (last 30 days)
4,000
Hugging Face likes
79
Listed
Nov 3, 2026

Popular in the same category

Resources in similar research fields

Related slides and articles

Ones that name this resource in their titles, and ones close to its description

Learn from slides & articles