microsoft / IRNet

An algorithm for cross-domain NL2SQL
MIT License
264 stars 81 forks source link

please help #56

Open Saber-shm opened 1 year ago

Saber-shm commented 1 year ago

i run python3 train.py and it runs for like 10 to 15 min and then gives me this error:

The size of tensor a (300) must match the size of tensor b (100) at non-singleton dimension 2 Traceback (most recent call last): File "/mnt/c/Users/Legion/irnet/IRNet/train_test.py", line 56, in train loss = utils.epoch_train(model, optimizer, args.batch_size, sql_data, table_data, args, File "/mnt/c/Users/Legion/irnet/IRNet/src/utils.py", line 242, in epoch_train score = model.forward(examples) File "/mnt/c/Users/Legion/irnet/IRNet/src/models/model.py", line 215, in forward table_embedding = table_embedding + tab_ctx RuntimeError: The size of tensor a (300) must match the size of tensor b (100) at non-singleton dimension 2

Monte2345 commented 1 year ago

the argument “hidden_size” should be set to 300?