Source:AISHELL-1: An Open-Source Mandarin Speech Corpus and A Speech Recognition Baseline Homepage Benchmarks Edit Add a new resultLink an existing benchmark TrendTaskDataset VariantBest ModelPaperCode Speech Recognition AISHELL-1 FireRedASR-AED ...
训练验证测试集 蛋 蛋宝哒哒 1枚 GPL 2 语音识别 8 41 2024-05-07 详情 相关项目 评论(0) 创建项目 文件列表 wav_-15_to_15db_test_4noise.zip avg_10.pdparams wav_-15_to_15db_train_dev_4noise.zip wav_-15_to_15db_test_4noise.zip (929.53M) 下载 File Name Size Update Time test/S...
WERs of 3.8%/9.1% on Librispeech test clean/other dataset without an external LM, and a CER of 5.8% on Aishell1 Mandarin corpus, respectively\n1\n... R Fan,W Chu,P Chang,... - IEEE International Conference on Acoustics 被引量: 0发表: 2021年 Improving Transformer-Based Speech Recogniti...
Prepare Dataset cd egs/aishell1 # Those stages are very time-consuming bash prepare.sh --stage -1 --stop-stage 3 ## train Cut statistics: ╒═══════════════════════════╤═══════════╕│ Cuts count: │ 120098 │├────────────...
$CUDA_VISIBLE_DEVICES=<gpu> python gvector_extraction.py <path-to-dataset-dir> --gvec_ckpt=<path-to-speaker-encoder-checkpoint> Train base synthesizer, first set the proper batch-size and gpu-numbers insynthesizer/hparams.py: # file: synthesizer/hprams.py tacotron_num_gpus = <n_gpus>, ...