From 291e27b39b78e2f29aca1b18c48888f86488d1b1 Mon Sep 17 00:00:00 2001 From: Mahesh Nandwana Date: Fri, 28 Jun 2024 17:03:54 -0700 Subject: [PATCH] Update README.md --- README.md | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/README.md b/README.md index 5620fee..5874736 100644 --- a/README.md +++ b/README.md @@ -1,6 +1,5 @@ ## Model Description -We present a large classification model trained on a manually curated real-world dataset that can be used as a new -benchmark for advancing research in toxicity detection and classification. +We present a large classification model trained on a manually curated real-world dataset that can be used as a new benchmark for advancing research in voice toxicity detection and classification. We started with the original weights from the [WavLM base plus](https://arxiv.org/abs/2110.13900) and fine-tuned it with 2,374 hours of voice chat audio clips for multilabel classification. The audio clips are automatically labeled using a synthetic data pipeline described in [our blog post](link to blog post here). A single output can have multiple labels. The model outputs a n by 6 output tensor where the inferred labels are `Profanity`, `DatingAndSexting`, `Racist`,