Where are model files going to?

Something seems wrong with how models are being stored in the database. Despite having run this model for nearly 3 epochs (over the entire weekend), I can't find it anywhere in the database. Any ideas @yamins81?

Since I passed no arguments regarding where to save everything should just be the defaults:

    Option                             Description                                                                  Default                 
    [--adp-drop <0/1>              ] - Adaptive Drop Training                                                       [False]                 
    [--check-grads <0/1>           ] - Check gradients and quit?                                                    [0]                     
    [--checkpoint-db-name <string> ] - Name for mongodb database for saved checkpoints                              [convnet_checkpoint_db] 
    [--checkpoint-fs-host <string> ] - Host for Saving Checkpoints to DB                                            [localhost]             
    [--checkpoint-fs-name <string> ] - Name for gridfs FS for saved checkpoints                                     [convnet_checkpoint_fs] 
    [--checkpoint-fs-port <int>    ] - Port for Saving Checkpoints to DB                                            [27017]

Here was the exact command I used.

python convnet.py --train-range=91-5039 --test-range=0-90 --layer-def=/home/ardila/src/archconvnets/archconvnets/convnet/nyu_model/Zeiler2013.cfg --layer-params=/home/ardila/src/archconvnets/archconvnets/convnet/nyu_model/layer-params.cfg --data-path=/export/storage2/ardila/small_challenge_batches --data-provider=general-cropped --test-freq=200 --conserve-mem=1 --max-filesize=99999999  --img-size=256 --save-db=1 --saving-freq=5 --experiment-data='{"experiment_id": "nyu_model_2"}' --dp-params='{"crop_border": 16, "meta_attribute": "synset", "preproc": {"normalize": false, "dtype": "float32", "resize_to": [256, 256], "mode": "RGB", "mask": null, "crop":null}, "batch_size": 256, "dataset_name": ["imagenet.dldatasets", "ChallengeSynsets2013_offline"]}'  --random-seed=1 --mini=256 & screen -d

dicarlolab / archconvnets

Where are model files going to? #27