Kassa
Kassa
The command is as follows: ``` python -m torch.distributed.launch --nproc_per_node 4 --master_port 12 train.py --seed 2 --cfg Salesforce/T5_3b_finetune_spider_with_cell_value.cfg --run_name T5_3b_finetune_spider_with_cell_value --logging_strategy steps --logging_first_step true --logging_steps 4 --evaluation_strategy steps --eval_steps 500...
The ckpt I chosen is the highest eval score during the training steps. As you can see, it is different from the test score.
They are still a little different. ![image](https://user-images.githubusercontent.com/30862458/168479534-1f97b8db-628f-4a20-8165-877a57318c46.png)
I check the evaluation and prediction json file, and find they are indeed different, no matter when do_train=False or num_train_epoch=0. The different sqls are like follows, just a few conditions...
Well, it is actually t5-large in this cfg file. I forget to change the file name.
![image](https://user-images.githubusercontent.com/30862458/169258929-c6ae4050-ae1c-47e4-8905-d9e6cbcebeb7.png) ![image](https://user-images.githubusercontent.com/30862458/169258987-13e77f10-4be0-4524-a55f-21a66cca4fd2.png)
hello, I install the new version of CoreNLP, but I still face this problem. could you please tell me why?
If I directly use the DDP package, it does not work?
I also want to know how to construct the tree using normal vector data
you can set scale_emb_or_prj=emb, and the result would be better