search

Plachtaa / VITS-fast-fine-tuning

This repo is a pipeline of VITS finetuning for fast speaker adaptation TTS, and many-to-many voice conversion
Apache License 2.0
4.69k stars 705 forks source link

训练时cuda超过memory了有办法解决吗 #464

Open light100bc opened 11 months ago

light100bc commented 11 months ago

运行finetune_speaker_v2.py (不是long_audio_transcribe.py，那里有自动改大小的选项)显示错误： torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 20.00 MiB (GPU 0; 15.70 GiB total capacity; 4.53 GiB already allocated; 18.12 MiB free; 4.69 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

有人知道代码改哪里能运行嘛。

tea9296 commented 11 months ago

In configs/finetune_speaker.json

Change "batch_size": 16 to smaller number