Update app.py

app.py

@@ -217,15 +217,14 @@ def consume_llm_api_updater(prompt):
 
     )
     return completion.choices[0].message.content    
-def consume_llm_api(prompt):
+def consume_llm_api(prompt,messages=None):
     
     client = Groq(
         api_key="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"
     )
-
-    completion = client.chat.completions.create(
-        
-        model="llama3-70b-8192",
+    if messages is not None:
+            messages= messages
+    else:
         messages=[
 
             {
@@ -233,17 +232,21 @@ def consume_llm_api(prompt):
                 "content": prompt
             },
         ],
+    completion = client.chat.completions.create(
+        
+        model="llama3-70b-8192",
+        messages=messages,
 
-
+        # temperature=1,
+        # max_completion_tokens=1024,
         top_p=1,
         stream=True,
-
+        # stop=None,
     )
-
+    # return completion.choices[0].message.content
     for chunk in completion:
         if chunk.choices[0].delta.content:
             yield chunk.choices[0].delta.content
-# @st.cache_resource
 # def load_model():
 #     pipeline_ = AutoPipelineForInpainting.from_pretrained("kandinsky-community/kandinsky-2-2-decoder-inpaint", torch_dtype=torch.float16).to("cuda")
 #     return pipeline_
@@ -409,6 +412,16 @@ with column2:
 
 
                 for index,prompts_ in enumerate(dictionary['every_prompt_with_val'][::-1]):
+                    messages = []
+
+                    for i_ in dictionary['every_prompt_with_val']:
+                        if "@working" in i_[-1]:
+                            messages.append({"role":"user","content":i_[0]})
+                            # messages.append({"role":"assistant","content":dictionary['every_prompt_with_val'][i_][1]})
+                        else:
+                            messages.append({"role":"user","content":i_[0]})
+                            messages.append({"role":"assistant","content":i_[1]})
+                            
                     if prompts_[-1]=="@working":
                         if index==0:
                             st.write(prompts_[0].split(send_prompt())[-1].upper() if send_prompt() in prompts_[0] else prompts_[0].upper())
@@ -416,12 +429,12 @@ with column2:
                             while(len(data_need)==0):
                                 if len(prompts_)==3:
                                     try:
-                                        data_need = st.write_stream(consume_llm_api(prompts_[1]))
+                                        data_need = st.write_stream(consume_llm_api(prompts_[1],messages))
                                     except:
                                         data_need = st.write_stream(consume_llm_api_conditional(prompts_[1]))
                                 else:
                                     try:
-                                        data_need=st.write_stream(consume_llm_api(prompts_[0]))
+                                        data_need=st.write_stream(consume_llm_api(prompts_[0],messages))
                                     except:
                                         data_need=st.write_stream(consume_llm_api_conditional(prompts_[0]))