Building a Production Greek-English Speech Recognizer
Engineering report details Sophea, a production Greek-English ASR reaching 4.26% WER on public English sets via ROVER ensemble and data-pipeline calibration.
Across 23 training iterations, two architectures, and nine production gates, no single data composition passed all gates; a three-model ROVER ensemble reached 9 of 9 gates and cut overlapping-speech WER from 53.35% to 37.87%. Calibrating an audio-quality filter against in-domain anchors reduced discarded scored Greek audio from 98.7% to 10.6%, and a pre-registered ablation traced a hallucination defect to one training-data package. The sophea/asr-k1 preview arbiter lists 4.26% average WER on eight public English test sets and 25.88% WER on live Greek noisy traffic; no weights or training data are released.
- ROVER ensemble of three models passed all 9 production gates; individual models passed 4-7.
- Audio-quality filter calibration cut discarded Greek audio from 98.7% to 10.6%.
- Pre-registered ablation isolated a hallucination defect to one training-data package.
- sophea/asr-k1 preview: 4.26% average WER on eight English test sets; 25.88% on Greek noisy traffic.
From the original article
1 min read at huggingface.co
We report a multi-month engineering program to build Sophea, a production bilingual Greek-English automatic speech recognition system. We evaluate the system against nine production gates covering Greek and English word error rate, language identification, and hallucinations on non-speech audio.
By Christos Petrocheilos, Cleopatra Papadopoulou, Chris Porikis · huggingface.co
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.