Update app.py

app.py

@@ -1,50 +1,39 @@
-# Step 1: Import Libraries
-import os
-import torch
-from transformers import AutoTokenizer, AutoModelForCausalLM
-import gradio as gr
-
-# Step 2: Load Environment Variables
-# Load token from environment variable
-HUGGING_FACE_TOKEN = os.getenv("HF_TOKEN")
-
-# Step 3: Load the Model with Authentication
-# Load Mistral 7B with authentication
-mistral_tokenizer = AutoTokenizer.from_pretrained(
-    "mistralai/Mistral-7B-v0.1",
-    use_auth_token=HUGGING_FACE_TOKEN
-)
-mistral_model = AutoModelForCausalLM.from_pretrained(
-    "mistralai/Mistral-7B-v0.1",
-    use_auth_token=HUGGING_FACE_TOKEN,
-    torch_dtype=torch.float16,  # Use half-precision for faster inference
-    device_map="auto"  # Automatically loads the model on GPU if available
-)
-
-# Step 4: Define the Chatbot Logic
-def generate_response(prompt):
-    # Tokenize the input prompt
-    inputs = mistral_tokenizer(prompt, return_tensors="pt").to(mistral_model.device)
-    
-    # Generate the response
-    outputs = mistral_model.generate(**inputs, max_length=200)
-    
-    # Decode the response
-    response = mistral_tokenizer.decode(outputs[0], skip_special_tokens=True)
-    return response
-
-# Step 5: Create the Gradio Interface
-def chatbot(prompt):
-    response = generate_response(prompt)
-    return response
-
-interface = gr.Interface(
-    fn=chatbot,
-    inputs="text",
-    outputs="text",
-    title="Mistral 7B Chatbot",
-    description="Ask questions and get answers from Mistral 7B!"
-)
-
-# Step 6: Launch the App
+from transformers import AutoTokenizer, AutoModelForCausalLM
+import torch
+
+# Load Zephyr 7B (no authentication required)
+zephyr_tokenizer = AutoTokenizer.from_pretrained("HuggingFaceH4/zephyr-7b-alpha")
+zephyr_model = AutoModelForCausalLM.from_pretrained(
+    "HuggingFaceH4/zephyr-7b-alpha",
+    torch_dtype=torch.float16,  # Use half-precision for faster inference
+    device_map="auto"  # Automatically loads the model on GPU if available
+)
+
+def generate_response(prompt):
+    # Tokenize the input prompt
+    inputs = zephyr_tokenizer(prompt, return_tensors="pt").to(zephyr_model.device)
+    
+    # Generate the response
+    outputs = zephyr_model.generate(**inputs, max_length=200)
+    
+    # Decode the response
+    response = zephyr_tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return response
+
+import gradio as gr
+
+# Gradio interface
+def chatbot(prompt):
+    response = generate_response(prompt)
+    return response
+
+interface = gr.Interface(
+    fn=chatbot,
+    inputs="text",
+    outputs="text",
+    title="Zephyr 7B Chatbot",
+    description="Ask questions and get answers from Zephyr 7B!"
+)
+
+# Launch the app
 interface.launch()