bi-att-flow
bi-att-flow copied to clipboard
allenai
FileNotFoundError: [Errno 2] No such file or directory: 'inter_single/eval.json'
This was previously closed but I am not sure why. The folder "inter_single/" exists with two .json files in it, neither of which are eval.json
Has anyone identified how to get a copy of this file?
@rjonesxmoz Have you figured where to get a copy of that file? I manually made a file 'eval.json' but I suppose it shouldn't be empty because I'm getting an out of input error. I'm getting the following error after running the script run_single.sh :--
(vur) srishti@VMtwo:~/notebooks/bi-att-flow-demo$ bash basic/run_single.sh $HOME /data/squad/dev-v1.1.json single.json
100%|███████████████████████████████████████████| 48/48 [00:04<00:00, 10.64it/s]
36%|█████████████████████████████████████████████████████▉ | 144729/400000 [00:10<00:19, 13174.72it/s]
20521/26469 of word vocab have corresponding vectors in ./glove.6B.100d.txt
36%|█████████████████████████████████████████████████████▉ | 144729/400000 [00:04<00:07, 34407.53it/s]
20594/23632 of word vocab have corresponding vectors in ./glove.6B.100d.txt
saving ...
1224
{ 'answer_dir': 'out/basic/00/answer',
'answer_func': 'linear',
'answer_path': '',
'batch_size': 1,
'c2q_att': True,
'char_count_th': 50,
'char_emb_size': 8,
'char_out_size': 100,
'char_vocab_size': 281,
'cluster': True,
'cpu_opt': True,
'data_dir': 'inter_single',
'data_filter': 'max',
'debug': False,
'decay': 0.9,
'device': '/cpu:0',
'device_type': 'gpu',
'dump_answer': False,
'dump_eval': True,
'dump_pickle': True,
'dynamic_att': False,
'eval': True,
'eval_dir': 'out/basic/00/eval',
'eval_path': 'inter_single/eval.json',
'eval_period': 1000,
'filter_heights': '5',
'finetune': False,
'forward_name': 'single',
'hidden_size': 100,
'highway': True,
'highway_num_layers': 2,
'init_lr': 0.5,
'input_keep_prob': 0.8,
'keep_prob': 0.8,
'known_if_glove': True,
'len_opt': True,
'load': True,
'load_ema': True,
'load_path': 'save/37/save',
'load_step': 0,
'log_dir': 'out/basic/00/log',
'log_period': 100,
'logit_func': 'tri_linear',
'lower_word': True,
'max_to_keep': 20,
'mode': 'forward',
'model_name': 'basic',
'new_emb_mat': array([[-0.10957 , -0.0044138, 0.137 , ..., -0.56343 , 0.45623 ,
-0.1637 ],
[ 0.2842 , 0.59621 , -0.58147 , ..., 0.37506 , 0.11688 ,
-0.1206 ],
[ 0.36211 , 0.26966 , -0.14845 , ..., -0.18546 , -1.5487 ,
-0.0041004],
...,
[-0.079885 , 0.7614 , -0.91762 , ..., -0.46399 , 0.48056 ,
-0.26194 ],
[-0.083848 , 0.61089 , -0.15195 , ..., -0.53627 , -0.2864 ,
0.66807 ],
[ 0.032177 , 0.61358 , 0.027951 , ..., -0.17203 , 0.5637 ,
0.32766 ]], dtype=float32),
'num_epochs': 12,
'num_gpus': 1,
'num_sents_th': 8,
'num_steps': 20000,
'out_base_dir': 'out',
'out_channel_dims': '100',
'out_dir': 'out/basic/00',
'para_size_th': 256,
'progress': True,
'q2c_att': True,
'ques_size_th': 30,
'run_id': '0',
'save_dir': 'out/basic/00/save',
'save_period': 1000,
'sent_size_th': 400,
'sh_logit_func': 'tri_linear',
'share_cnn_weights': True,
'share_lstm_weights': True,
'shared_path': 'save/37/shared.json',
'single': False,
'squash': False,
'swap_memory': True,
'test_num_batches': 0,
'use_char_emb': True,
'use_glove_for_unk': True,
'use_word_emb': True,
'val_num_batches': 100,
'var_decay': 0.999,
'vis': False,
'wd': 0.0,
'word_count_th': 10,
'word_emb_size': 100,
'word_size_th': 16,
'word_vocab_size': 1224}
Traceback (most recent call last):
File "/anaconda/envs/vur/lib/python3.5/runpy.py", line 184, in _run_module_as_main
"main", mod_spec)
File "/anaconda/envs/vur/lib/python3.5/runpy.py", line 85, in _run_code
exec(code, run_globals)
File "/home/srishti/notebooks/bi-att-flow-demo/basic/cli.py", line 109, in
File "/anaconda/envs/vur/lib/python3.5/site-packages/tensorflow/python/platform/flags.py", line 50, in getattr
raise AttributeError(name)
AttributeError: max_num_sents
0%| | 0/1 [00:00<?, ?it/s]
Traceback (most recent call last):
File "/anaconda/envs/vur/lib/python3.5/runpy.py", line 184, in _run_module_as_main
"main", mod_spec)
File "/anaconda/envs/vur/lib/python3.5/runpy.py", line 85, in _run_code
exec(code, run_globals)
File "/home/srishti/notebooks/bi-att-flow-demo/basic/ensemble.py", line 106, in
EOFError: Ran out of input
@rjonesxmoz @houzhenzhen @srishteeh
The eval.json in directory inter_single is generated after running the process.
I run the command in run_single.sh one by one. After I running the command: (just replace the parameters with the absolute path) python3 -m basic.cli --data_dir inter_single --eval_path inter_single/eval.json --nodump_answer --load_path save/37/save --shared_path save/37/shared.json "" --eval_num_batches 0 --mode forward --batch_size 1 --len_opt --cluster --cpu_opt --load_ema
The file inter_single/eval.json is generated, and it looks like a file wrote in binary form. By the way, the environment is import. My python version is 3.5.2, and the tensorflow version is 0.12.0.
I found an eval.json file in this bundle. It is in the inter_single folder.
However I don't understand why this file is needed during evaluation.
Also I found that such files are saved during training in out/basic/00/eval/.
