Make cached_token an option that can change; and lazy intialize the c…

…ider scorer.

misc/rewards.py

@@ -13,9 +13,13 @@
 sys.path.append("cider")
 from pyciderevalcap.ciderD.ciderD import CiderD
 
-CiderD_scorer = CiderD(df='coco-train-idxs')
+CiderD_scorer = None
 #CiderD_scorer = CiderD(df='corpus')
 
+def init_cider_scorer(cached_tokens):
+    global CiderD_scorer
+    CiderD_scorer = CiderD(df=cached_tokens)
+
 def array_to_str(arr):
     out = ''
     for i in range(len(arr)):

opts.py

@@ -18,6 +18,8 @@ def parse_opt():
                                               Note: this file contains absolute paths, be careful when moving files around;
                         'model.ckpt-*'      : file(s) with model definition (created by tf)
                     """)
+    parser.add_argument('--cached_tokens', type=str, default='coco-train-idxs',
+                    help='Cached token file for calculating cider score during self critical training.')
 
     # Model settings
     parser.add_argument('--caption_model', type=str, default="show_tell",

train.py

@@ -18,7 +18,7 @@
 from dataloader import *
 import eval_utils
 import misc.utils as utils
-from misc.rewards import get_self_critical_reward
+from misc.rewards import init_cider_scorer, get_self_critical_reward
 
 try:
     import tensorflow as tf
@@ -101,6 +101,7 @@ def train(opt):
             # If start self critical training
             if opt.self_critical_after != -1 and epoch >= opt.self_critical_after:
                 sc_flag = True
+                init_cider_scorer(opt.cached_tokens)
             else:
                 sc_flag = False