Merge pull request #1 from yashkant/update-py3

Update to Python3 and Rip-Off Lua code

.gitignore

@@ -9,3 +9,12 @@ ques_feat.json
 models/*.caffemodel
 models/*.lua
 models/*.prototxt
+*.zip
+
+# Pycharm
+.idea/
+
+# Installed packages
+pytorch/
+migrations/
+!migrations/__init__.py

README.md

@@ -139,6 +139,10 @@ If you find this code useful, consider citing our work:
 
 BSD
 
+## Helpful Issues 
+Problems installing uwsgi: https://github.com/unbit/uwsgi/issues/1770 
+
+Problems with asgiref: https://stackoverflow.com/questions/41335478/importerror-no-module-named-asgiref-base-layer 
 ## Credits
 
 - Visual Chatbot Image: "[Robot-clip-art-book-covers-feJCV3-clipart](https://commons.wikimedia.org/wiki/File:Robot-clip-art-book-covers-feJCV3-clipart.png)" by [Wikimedia Commons](https://commons.wikimedia.org) is licensed under [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/deed.en)

chat/consumers.py

@@ -2,11 +2,11 @@
 from chat.utils import log_to_terminal
 
 def ws_connect(message):
-    print "User connnected via Socket"
+    print("User connnected via Socket")
 
 
 def ws_message(message):
-    print "Message recieved from client side and the content is ", message.content['text']
+    print("Message recieved from client side and the content is ", message.content['text'])
     # prefix, label = message['path'].strip('/').split('/')
     socketid = message.content['text']
 

chat/views.py

@@ -13,7 +13,7 @@
 import os
 import traceback
 import random
-import urllib2
+import urllib
 
 
 def home(request, template_name="chat/index.html"):
@@ -28,7 +28,7 @@ def home(request, template_name="chat/index.html"):
             job_id = request.POST.get("job_id")
             history = request.POST.get("history", "")
 
-            img_path = urllib2.unquote(img_path)
+            img_path = urllib.parse.unquote(img_path)
             abs_image_path = str(img_path)
 
             q_tokens = word_tokenize(str(question))
@@ -38,7 +38,7 @@ def home(request, template_name="chat/index.html"):
             response = svqa(str(question), str(history),
                             str(abs_image_path), socketid, job_id)
             return JsonResponse({"success": True})
-        except Exception, err:
+        except Exception:
             return JsonResponse({"success": False})
 
     elif request.method == "GET":

models.py

@@ -0,0 +1,57 @@
+class VisDialDummyModel:
+
+    def __init__(self, inputJson, loadPath, beamSize, beamLen, sampleWords,
+                 temperature, gpuid, backend, proto_file, model_file,
+                 maxThreads, encoder, decoder):
+        print("init visdial model")
+        self.decoder = decoder
+        self.encoder = encoder
+        self.maxThreads = maxThreads
+        self.proto_file = proto_file
+        self.model_file = model_file
+        self.backend = backend
+        self.gpuid = gpuid
+        self.sampleWords = sampleWords
+        self.beamLen = beamLen
+        self.beamSize = beamSize
+        self.loadPath = loadPath
+        self.inputJson = inputJson
+        self.temperature = temperature
+
+    def predict(self, img, history, question):
+        print("predict-visdial called!")
+        print("img: ", img)
+        print("hist: ", history)
+        print("ques: ", question)
+        dummy_ans_str = "dummy ans here to your dummy question there"
+        dummy_hist_str = question + " " + dummy_ans_str
+
+        result = {'answer': dummy_ans_str,
+                  'question': question,
+                  'history': ''.join(history) + dummy_hist_str, 'input_image': img}
+        print(result)
+        return result
+
+
+class CaptioningTorchDummyModel:
+
+    def __init__(self, model_path, backend, input_sz, layer, seed, gpuid):
+        print("init caption model")
+        self.seed = seed
+        self.layer = layer
+        self.input_sz = input_sz
+        self.model_path = model_path
+        self.backend = backend
+        self.gpuid = gpuid
+        self.loadModel(model_path)
+
+    def loadModel(self, model_path):
+        print("load-model called!")
+        print("model_path: ", model_path)
+
+    def predict(self, input_image_path, input_sz1, input_sz2):
+        print("predict-caption called!")
+        dummy_caption_str = "dummy caption here"
+        result = {'input_image': input_image_path,
+                  'pred_caption': dummy_caption_str}
+        return result

worker.py

@@ -9,18 +9,14 @@
 from django.conf import settings
 from chat.utils import log_to_terminal
 from chat.models import Job, Dialog
-
+from models import VisDialDummyModel
 import chat.constants as constants
-import PyTorch
-import PyTorchHelpers
 import pika
-import time
 import yaml
 import json
 import traceback
 
-VisDialModel = PyTorchHelpers.load_lua_class(
-    constants.VISDIAL_LUA_PATH, 'VisDialTorchModel')
+VisDialModel = VisDialDummyModel
 
 VisDialATorchModel = VisDialModel(
     constants.VISDIAL_CONFIG['input_json'],
@@ -47,6 +43,7 @@
 
 django.db.close_old_connections()
 
+
 def callback(ch, method, properties, body):
     try:
         body = yaml.safe_load(body)
@@ -69,12 +66,12 @@ def callback(ch, method, properties, body):
             job = Job.objects.get(id=int(body['job_id']))
             Dialog.objects.create(job=job, question=result['question'], answer=result['answer'].replace("<START>", "").replace("<END>", ""))
         except:
-            print str(traceback.print_exc())
+            print(str(traceback.print_exc()))
 
         django.db.close_old_connections()
 
-    except Exception, err:
-        print str(traceback.print_exc())
+    except Exception:
+        print(str(traceback.print_exc()))
 
 channel.basic_consume(callback,
                       queue='visdial_task_queue')

worker_captioning.py

@@ -8,22 +8,21 @@
 
 from django.conf import settings
 from chat.utils import log_to_terminal
-from chat.models import Job, Dialog
+from chat.models import Job
 
 import chat.constants as constants
 
-import PyTorch
-import PyTorchHelpers
 import pika
-import time
 import yaml
 import json
 import traceback
 
+from models import CaptioningTorchDummyModel
+
 django.db.close_old_connections()
 
-CaptioningModel = PyTorchHelpers.load_lua_class(
-    constants.CAPTIONING_LUA_PATH, 'CaptioningTorchModel')
+CaptioningModel = CaptioningTorchDummyModel
+
 CaptioningTorchModel = CaptioningModel(
     constants.CAPTIONING_CONFIG['model_path'],
     constants.CAPTIONING_CONFIG['backend'],
@@ -52,16 +51,17 @@ def callback(ch, method, properties, body):
         result['input_image'] = str(result['input_image']).replace(settings.BASE_DIR, '')
         log_to_terminal(body['socketid'], {"result": json.dumps(result)})
         ch.basic_ack(delivery_tag=method.delivery_tag)
+        print('succesfull callback')
 
         try:
             Job.objects.filter(id=int(body['job_id'])).update(caption=result['pred_caption'])
         except Exception as e:
-            print str(traceback.print_exc())
+            print(str(traceback.print_exc()))
 
         django.db.close_old_connections()
 
-    except Exception, err:
-        print str(traceback.print_exc())
+    except Exception:
+        print(str(traceback.print_exc()))
 
 channel.basic_consume(callback,
                       queue='visdial_captioning_task_queue')