lstm gru VAE
lstm based PyTorch implementation for lora pretrained.
- Input
- 1891-dim embedding
- Encoder
- 53 x VAE with 28 heads
- Output
- auc-roc projection
Training config
optimizer=SGD, lr=0.258, scheduler=linear, warmup=82无论您有任何问题或需求,我们的团队随时准备为您服务
复制添加QQ,立即沟通