pcyin / tranX

A general-purpose neural semantic parser for mapping natural language queries into machine executable code
Apache License 2.0
460 stars 111 forks source link

Need scripts directory to train #4

Open rajasagashe opened 5 years ago

rajasagashe commented 5 years ago

The Readme says: "The scripts folder contains scripts to train TranX on example datasets. For example, to train on the Django dataset, simply run: . scripts/django/train.sh 0 "

When will this be added?

pcyin commented 5 years ago

Hi! Please run rull_data.sh to download the example datasets and training scripts

armheb commented 5 years ago

after running rull_data.sh still can't train the django model, it finishes without an error but doesn't seem to do anything. what should I do?

gsh2014 commented 5 years ago

same situation,have you solved this question

after running rull_data.sh still can't train the django model, it finishes without an error but doesn't seem to do anything. what should I do?

pcyin commented 5 years ago

Hi all, sorry for the late reply! Could you try manually downloading http://www.cs.cmu.edu/~pengchey/tranx.0.2.0.zip and unzip it? The scripts folder is in the zip file :)

gsh2014 commented 5 years ago

I have got the scripts,but as @armheb said it finished without an error,after I run the command "./scripts/atis/train.sh 0" . I think the scripts may have some problems. :)

pcyin commented 5 years ago

@gsh2014 Could you post your error message? Meanwhile, please make sure you use the correct conda environment to run the training script. django dataset only supports python 2.x

gsh2014 commented 5 years ago

there is no error message at all , maybe the " python -u exp.py ..." in the "./scripts/atis/train.sh 0 " hasn't been excuted for some reason .

pcyin commented 5 years ago

This's weird. Could you check if logs/django/${model_name}.log exists, and if so, what's the content? Also, make sure the folder logs/django exists. pull_data.sh should generate it for you. You may also try commenting out the redirection, and post the standard output/err messages here.

gsh2014 commented 5 years ago

Namespace(action_embed_size=128, answer_prune=True, asdl_file='asdl/lang/lambda_dcs/lambda_asdl.txt', att_vec_size=256, batch_size=10, beam_size=5, clip_grad=5.0, column_att='affine', cuda=True, decay_lr_every_epoch=False, decode_max_time_step=110, decoder_word_dropout=0.3, dev_file='data/atis/dev.bin', dropout=0.3, embed_size=128, eval_top_pred_only=False, evaluator='default_evaluator', field_embed_size=32, glorot_init=True, glove_embed_path=None, hidden_size=256, lang='python', load_model=None, log_every=50, lr=0.001, lr_decay=0.5, lr_decay_after_epoch=0, lstm='lstm', max_epoch=-1, max_num_trial=5, mode='train', negative_sample_type='best', no_copy=True, no_input_feed=False, no_parent_field_embed=False, no_parent_field_type_embed=True, no_parent_production_embed=True, no_parent_state=False, no_query_vec_to_action_map=False, optimizer='Adam', parser='default_parser', patience=5, primitive_token_label_smoothing=0.1, ptrnet_hidden_dim=32, query_vec_to_action_diff_map=False, readout='identity', reset_optimizer=False, sample_size=5, save_all_models=False, save_decode_to=None, save_to='saved_models/atis/model.atis.sup.lstm.hidden256.embed128.action128.field32.type32.dropout0.3.lr_decay0.5.beam5.vocab.freq2.bin.train.bin.glorot.with_par_info.no_copy.ls0.1.seed0', seed=0, sql_db_file=None, src_token_label_smoothing=0.0, sup_attention=False, test_file=None, train_file='data/atis/train.bin', transition_system='lambda_dcs', type_embed_size=32, uniform_init=None, valid_every_epoch=1, valid_metric='acc', verbose=False, vocab='data/atis/vocab.freq2.bin', word_dropout=0.0) Traceback (most recent call last): File "exp.py", line 251, in <module> train(args) File "exp.py", line 71, in train if args.cuda: model.cuda() File "/home/guishuo/anaconda3/envs/py3torch3cuda9/lib/python3.6/site-packages/torch/nn/modules/module.py", line 216, in cuda return self._apply(lambda t: t.cuda(device)) File "/home/guishuo/anaconda3/envs/py3torch3cuda9/lib/python3.6/site-packages/torch/nn/modules/module.py", line 146, in _apply module._apply(fn) File "/home/guishuo/anaconda3/envs/py3torch3cuda9/lib/python3.6/site-packages/torch/nn/modules/rnn.py", line 123, in _apply self.flatten_parameters() File "/home/guishuo/anaconda3/envs/py3torch3cuda9/lib/python3.6/site-packages/torch/nn/modules/rnn.py", line 102, in flatten_parameters fn.rnn_desc = rnn.init_rnn_descriptor(fn, handle) File "/home/guishuo/anaconda3/envs/py3torch3cuda9/lib/python3.6/site-packages/torch/backends/cudnn/rnn.py", line 42, in init_rnn_descriptor cudnn.DropoutDescriptor(handle, dropout_p, fn.dropout_seed) File "/home/guishuo/anaconda3/envs/py3torch3cuda9/lib/python3.6/site-packages/torch/backends/cudnn/__init__.py", line 207, in __init__ self._set(dropout, seed) File "/home/guishuo/anaconda3/envs/py3torch3cuda9/lib/python3.6/site-packages/torch/backends/cudnn/__init__.py", line 232, in _set ctypes.c_ulonglong(seed), File "/home/guishuo/anaconda3/envs/py3torch3cuda9/lib/python3.6/site-packages/torch/backends/cudnn/__init__.py", line 283, in check_error raise CuDNNError(status) torch.backends.cudnn.CuDNNError: 8: b'CUDNN_STATUS_EXECUTION_FAILED'

pcyin commented 5 years ago

Hi @gsh2014, this might be a cuda-related issue, could you try updating your nvidia driver? Also, you could also try running the code on CPU without the --cuda flag.

pcyin commented 5 years ago

I am closing the issue since it might be platform/driver-related. Please feel free to re-open it!

gsh2014 commented 5 years ago

thanks for your patience :)