feat/streaming (#18)

* feat/streaming

streaming endpoint, faster responses!

* Update requirements.txt

* Update __init__.py

__init__.py

@@ -13,13 +13,8 @@ def runtime_requirements(self):
         return RuntimeRequirements(
             internet_before_load=True,
             network_before_load=True,
-            gui_before_load=False,
             requires_internet=True,
-            requires_network=True,
-            requires_gui=False,
-            no_internet_fallback=False,
-            no_network_fallback=False,
-            no_gui_fallback=True,
+            requires_network=True
         )
 
     def initialize(self):
@@ -75,11 +70,15 @@ def build_msg_history(self, message):
     def _async_ask(self, message):
         utterance = message.data["utterance"]
         self.chat.qa_pairs = self.build_msg_history(message)
-        answer = self.chat.get_spoken_answer(utterance)
-        if not answer:
+        answered = False
+        try:
+            for utt in self.chat.stream_utterances(utterance):
+                answered = True
+                self.speak(utt)
+        except: # speak error on any network issue / no credits etc
+            pass
+        if not answered:
             self.speak_dialog("gpt_error")
-        else:
-            self.speak(answer)
 
     def ask_chatgpt(self, message):
         if "key" not in self.settings:

requirements.txt

@@ -1,5 +1,5 @@
-ovos-solver-openai-persona-plugin
+ovos-solver-openai-persona-plugin>=0.0.0a8
 ovos-utils~=0.0, >=0.0.28
 ovos_workshop~=0.0, >=0.0.11
-ovos-plugin-manager~=0.0, >=0.0.23a13
-ovos-translate-server-plugin
+ovos-plugin-manager~=0.0, >=0.0.26a9
+ovos-translate-server-plugin