Update chain_app.py

chain_app.py

@@ -6,6 +6,8 @@ import requests
 from chainlit.input_widget import Select, Slider
 import os
 import cohere 
+from huggingface_hub import InferenceClient
+
 
 hf_token = os.environ.get("HF_TOKEN")
 openai_api_key = os.environ.get('OPENAI_API_KEY')
@@ -77,9 +79,21 @@ async def chat_profile():
             name="TTS",
             markdown_description="OpenAI's Text-to-Speech model",
         ),
+        cl.ChatProfile(
+            name="Llama-3.1-405B",
+            markdown_description="Meta Open Source Model Llama with 405B parameters",
+        ),
+        cl.ChatProfile(
+            name="Llama-3.1-70B",
+            markdown_description="Meta Open Source Model Llama with 70B parameters",
+        ),
+        cl.ChatProfile(
+            name="Llama-3.1-8B",
+            markdown_description="Meta Open Source Model Llama with 8B parameters",
+        ),
         cl.ChatProfile(
             name="Llama-3-70B",
-            markdown_description="Meta Open Source model Llama-2 with 70B parameters",
+            markdown_description="Meta Open Source model Llama-3 with 70B parameters",
         ),
         cl.ChatProfile(
             name='Aya-23B',
@@ -312,6 +326,30 @@ async def on_chat_start():
         await cl.Message(
             content="Im TTS. of the best models OpenAI ever created. i can convert text to speech! . i was configured by Artin Daneshvar and Sadra Noadoust, 2 iranian students to help you, how can i assist you today ? "
         ).send()
+
+    if chat_profile == 'Llama-3.1-405B':
+        await cl.ChatSettings(
+            [
+                Select(
+                    id="Meta-Model",
+                    label="Meta - Model",
+                    values=["Llama-3-70B"],
+                    initial_index=0,
+                ),
+                Slider(
+                    id="Temperature",
+                    label="Model Temperature",
+                    initial=0.7,
+                    min=0,
+                    max=1,
+                    step=0.1,
+                ),
+            ]
+        ).send()
+        await cl.Message(
+            content="Im the big Llama-3.1!. one of the best open source models released by Meta! i am the Big version of meta's open source LLMs., i was configured by Artin Daneshvar and Sadra Noadoust, 2 iranian students to help you, how can i assist you today ? "
+        ).send()
+        
     if chat_profile == 'Llama-3-70B':
         await cl.ChatSettings(
             [
@@ -332,7 +370,7 @@ async def on_chat_start():
             ]
         ).send()
         await cl.Message(
-            content="Im the big Llama!. one of the best open source models released by Meta! i am the Big version of meta's open source LLMs., i was configured by Artin Daneshvar and Sadra Noadoust, 2 iranian students to help you, how can i assist you today ? "
+            content="Im the big Llama-3!. one of the best open source models released by Meta! i am the Big version of meta's open source LLMs., i was configured by Artin Daneshvar and Sadra Noadoust, 2 iranian students to help you, how can i assist you today ? "
         ).send()
     if chat_profile == 'Llama-3-8B':
         await cl.ChatSettings(
@@ -587,6 +625,23 @@ async def main(message: cl.Message):
             elements=elements,
         ).send()
 
+    elif chat_profile == 'Llama-3.1-405B':        
+        client = InferenceClient(
+            "meta-llama/Meta-Llama-3.1-405B-Instruct",
+            token=hf_token,
+        )
+        
+        for message in client.chat_completion(
+            messages=[{"role": "user", "content": f'{message.content}'}],
+            max_tokens=500,
+            stream=True,
+        ):
+            complete_message += message.choiches[0].delta.content
+            await cl.Message(
+                content=complete_message,
+            ).send()
+
+
     elif chat_profile == 'Llama-3-70B':
         completion = groq_client.chat.completions.create(
             model="llama3-70b-8192",