Xu Ding

Results 16 comments of Xu Ding
trafficstars

这个命令我执行过了,截图如下,是不是表示安装好了已经? ![image](https://user-images.githubusercontent.com/38311101/198792582-ea2e5e10-26d3-4369-81bd-44c05336baf6.png)

> thanks for your reply. so you mean the default operator is using FP16 format to train a new model? where can I find this setting code? very thanks!

@pacman100 thank you! I use the LoRA method to train my chatGLM models, below are the params: trainable params: 1572864 || all params: 1723981824 || trainable%: 0.09123437254985815 trainable params: 1572864...

LoRA config: LORA_PARA = { 'r': 8, 'lora_alpha': 32, 'lora_dropout': 0.05, 'target_modules': ['query_key_value'], 'task_type': "CAUSAL_LM", 'inference_mode': False, }