laixiaofeng0
Results
3
comments of
laixiaofeng0
@Kevinarcsin001 有解决了该问题嘛?我也碰到了相同的问题。
我尝试使用了8台40GB的A100,微调了aishell1和其它大量的数据,只在v3上微调有相似问题,在v2上没有复现到该问题。一直以为是batch数过少的原因导致的,设置per_device_train_batch_size=3,gradient_accumulation_steps=5,依然有相似的问题。
@niedev Hello, do you have any plans to release the Whisper model converted to ONNX? I’m very eager to know. Thank you.