tsukumi 15af441cba Improve: Convert ONNX version of BERT models to FP16 ...

I have found that half-precision has little effect on speech synthesis quality and, depending on the environment, can reduce file size and memory usage by half, so I have decided to use FP16.

2024-12-19 14:51:30 +09:00

..

chinese-roberta-wwm-ext-large

Fix: The conversion script for English BERT to Fast Tokenizer was incorrect, so tokenization was not performed correctly

2024-12-19 11:10:33 +09:00

chinese-roberta-wwm-ext-large-onnx

Fix: The conversion script for English BERT to Fast Tokenizer was incorrect, so tokenization was not performed correctly

2024-12-19 11:10:33 +09:00

deberta-v2-large-japanese-char-wwm

Add: Support for speech synthesis in English and Chinese for ONNX inference in Non-JP-Extra models

2024-09-23 07:14:44 +09:00

deberta-v2-large-japanese-char-wwm-onnx

Improve: Support ONNX inference, add ONNX conversion script

2024-09-17 16:55:03 +09:00

deberta-v3-large

Fix: The conversion script for English BERT to Fast Tokenizer was incorrect, so tokenization was not performed correctly

2024-12-19 11:10:33 +09:00

deberta-v3-large-onnx

Fix: The conversion script for English BERT to Fast Tokenizer was incorrect, so tokenization was not performed correctly

2024-12-19 11:10:33 +09:00

bert_models.json

Improve: Convert ONNX version of BERT models to FP16

2024-12-19 14:51:30 +09:00