f

app.py

@@ -1,36 +1,37 @@
 from transformers import AutoModelForCausalLM, AutoTokenizer
-import gradio as gr
 
-# Define the model paths
+# Specify the model paths
 base_model_name = "deepseek-ai/DeepSeek-R1-Distill-Qwen-7B"
 fine_tuned_model_name = "cheberle/autotrain-35swc-b4r9z"
 
 # Load the tokenizer
-tokenizer = AutoTokenizer.from_pretrained(fine_tuned_model_name)
+tokenizer = AutoTokenizer.from_pretrained(base_model_name, trust_remote_code=True)
 
-# Load the model
+# Load the base model with fine-tuned weights
 model = AutoModelForCausalLM.from_pretrained(
     fine_tuned_model_name,
-    device_map="auto",  # Auto-distributes model across devices
-    torch_dtype="auto", # Matches model precision
+    device_map="auto",
+    torch_dtype="auto",
+    trust_remote_code=True
 )
 
-# Define the chat function
+# Define a simple function for chat
 def chat(input_text):
-    input_ids = tokenizer(input_text, return_tensors="pt").input_ids.to("cuda")  # Move input to GPU
-    output = model.generate(input_ids, max_length=100)
-    response = tokenizer.decode(output[0], skip_special_tokens=True)
+    inputs = tokenizer(input_text, return_tensors="pt").input_ids.to("cuda")
+    outputs = model.generate(inputs, max_length=100, temperature=0.7)
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
     return response
 
-# Create a Gradio interface
+# Gradio UI
+import gradio as gr
+
 interface = gr.Interface(
     fn=chat,
-    inputs=gr.Textbox(lines=2, placeholder="Type your input here..."),
+    inputs=gr.Textbox(lines=2, placeholder="Type your message here..."),
     outputs="text",
-    title="Chat with DeepSeek-AutoTrain Model",
-    description="Fine-tuned version of DeepSeek-R1-Distill-Qwen-7B. Ask me anything!",
+    title="Chat with DeepSeek Fine-tuned Model",
+    description="This is a fine-tuned version of the DeepSeek R1 Distill Qwen-7B model. Ask me anything!"
 )
 
-# Launch the interface
 if __name__ == "__main__":
     interface.launch()