Update app.py

app.py

@@ -249,13 +249,17 @@ def process_audio_from_video(video_path):
 
 
 
-
+import torch
 import gradio as gr
 from huggingface_hub import InferenceClient
 from transformers import AutoTokenizer, AutoModelForCausalLM
-from huggingface_hub import InferenceClient
 
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+# Hugging Face Inference Client (equivalent to the reference code's client)
+client = InferenceClient("TheBloke/Mistral-7B-Instruct-v0.1-GPTQ")
+
+# Tokenizer and model loading (still necessary if you want to process locally)
+tokenizer = AutoTokenizer.from_pretrained("TheBloke/Mistral-7B-Instruct-v0.1-GPTQ")
+model = AutoModelForCausalLM.from_pretrained("TheBloke/Mistral-7B-Instruct-v0.1-GPTQ")
 
 
 def respond(message, history: list[tuple[str, str]], system_message, max_tokens, temperature, top_p):
@@ -286,10 +290,10 @@ def respond(message, history: list[tuple[str, str]], system_message, max_tokens,
 
 
 # Function to handle video processing and interaction
-def transcribe_and_predict_video(video, chat_history=[]):
+def transcribe_and_predict_video(video, user_input, chat_history=[]):
     # Process the video for emotions (use your own emotion detection functions)
     image_emotion = process_video(video)
-    text_emotion, audio_emotion, user_input = process_audio_from_video(video)
+    text_emotion, audio_emotion = process_audio_from_video(video)
     em = [image_emotion, text_emotion, audio_emotion]
 
     # Format the conversation history
@@ -324,20 +328,17 @@ def transcribe_and_predict_video(video, chat_history=[]):
 
 
 # Gradio interface setup
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
-    ],
+iface = gr.Interface(
+    fn=transcribe_and_predict_video,
+    inputs=[gr.Video(), gr.Textbox(), gr.State()],  # Accepting video input, user text, and chat history
+    outputs=[gr.Textbox(), gr.State()],  # Output is the response and updated chat history
+    title="Multimodal Emotion Recognition from Video",
+    description="Upload a video to get text, audio, and image emotion predictions and interact with the chatbot."
 )
 
 # Launch the Gradio interface
 if __name__ == "__main__":
-    demo.launch()
-
+    iface.launch()