Issue in the code to run the Train script and evaluation script about gar HOT 3 CLOSED

@hmhm1190 Following is an example command I used for the Spider dataset, FYI.

bash train_pipeline.sh "spider" "datasets/spider/train_spider.json" "datasets/spider/dev.json" "datasets/spider/tables.json" "datasets/spider/database"
bash test_pipeline.sh "spider" "datasets/spider/dev.json" "datasets/spider/dev_gold.sql" "datasets/spider/tables.json" "datasets/spider/database"

from gar.

Hi @Kaimary ,
Thanks so much for your response! Would help me with the necessary changes to test the model on my own database and NL Questions.

from gar.

@hmhm1190 To test the model on your own database, you just need to prepare the sample query file and schema file(Spider data format) and run the test_pipeline.sh as above. If you have a separate sample query file and test NL query file, please use the generalization_script.py script to do the generalization first, and then run the test_pipeline.sh script.

from gar.