LUMIA-Group/rasat
The official implementation of the paper "RASAT: Integrating Relational Structures into Pretrained Seq2Seq Model for Text-to-SQL"(EMNLP 2022)
PythonApache-2.0
Issues
- 0
- 0
the version of transformers
#26 opened by lllxf - 0
Eval process issue
#25 opened by fengxin-zhxx - 1
Not able to clone the repo
#23 opened by Tizzzzy - 1
Training process bug
#24 opened by bekhzod-olimov - 0
problem with train/eval
#22 opened by SepehrAminiAfshar - 0
Error Message "size mismatch for relation_k_emb.weight" when i'm trying load a training models using t5-small
#21 opened by kanseaveg - 2
Hello, I have a small question for you. What is the meaning of the sequence number after the model configuration file?
#20 opened by kanseaveg - 3
share the predicted SQL file
#18 opened by linxin6 - 9
- 1
Picard Client Launch Issue
#16 opened by Rautmare-Sneha - 1
TypeError: issubclass() arg 1 must be a class
#19 opened by prattsoroco - 4
Model weight missing.
#17 opened by linxin6 - 1
Possible error in `generate_relations`
#14 opened by FruVirus - 1
- 2
Why is table alignment necessary?
#13 opened by FruVirus - 0
Indexing error if max_train_samples is not 7000
#12 opened by FruVirus - 5
- 1
- 1
- 1
How make prediction on trained model ?
#7 opened by pmane-uptycs - 1
cspider数据集需要怎么训练呢?
#5 opened by szj2ys - 1
- 1
model_cls_wrapper is not used
#6 opened by Dongfeng-He - 4
Could you share the training config of sparc_add_coref_t5_3b_order_0514_ckpt-4224?
#3 opened by Dongfeng-He - 1
- 2
File missing when starting training
#1 opened by OverRipeThree49