Merge pull request #117 from skysign/master

Transferring 'current_seq' to CPU from GPU

Author: cezannec

project-tv-script-generation/dlnd_tv_script_generation.ipynb

@@ -786,6 +786,8 @@
     "        word = int_to_vocab[word_i]\n",
     "        predicted.append(word)     \n",
     "        \n",
+    "        if(train_on_gpu):\n",
+    "            current_seq = current_seq.cpu() # move to cpu\n",
     "        # the generated word becomes the next \"current sequence\" and the cycle can continue\n",
     "        if train_on_gpu:\n",
     "            current_seq = current_seq.cpu()\n",