Update app.py

app.py

@@ -1,26 +1,29 @@
-import gradio as gr
-from transformers import AutoModelForCausalLM, AutoTokenizer
-
-model_name = "deepseek-ai/deepseek-coder-1.3b-instruct"
-tokenizer = AutoTokenizer.from_pretrained(model_name, use_auth_token=True)
-model = AutoModelForCausalLM.from_pretrained(model_name, use_auth_token=True)
-
-def respond(query):
-    prompt = f"[INST] {query} [/INST]"
-    inputs = tokenizer(prompt, return_tensors="pt")
-    outputs = model.generate(
-        **inputs,
-        max_new_tokens=250,
-        pad_token_id=tokenizer.eos_token_id
-    )
-    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
-    return response
-
-with gr.Blocks() as demo:
-    gr.Markdown("# Deepseek Coder Chatbot")
-    query_input = gr.Textbox(label="Ask me anything...")
-    output = gr.Textbox(label="Response")
-    submit_button = gr.Button("Submit")
-    submit_button.click(respond, inputs=query_input, outputs=output)
-
+import gradio as gr
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+# Load the model and tokenizer
+model_name = "deepseek-ai/deepseek-coder-1.3b-instruct"
+tokenizer = AutoTokenizer.from_pretrained(model_name, use_auth_token=True)
+model = AutoModelForCausalLM.from_pretrained(model_name, use_auth_token=True)
+
+# Define the response function
+def respond(query):
+    prompt = f"[INST] {query} [/INST]"
+    inputs = tokenizer(prompt, return_tensors="pt")
+    outputs = model.generate(
+        **inputs,
+        max_new_tokens=50,  # Adjust based on resource constraints
+        pad_token_id=tokenizer.eos_token_id
+    )
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return response
+
+# Create the Gradio interface
+with gr.Blocks() as demo:
+    gr.Markdown("# Deepseek Coder Chatbot")
+    query_input = gr.Textbox(label="Ask me anything...")
+    output = gr.Textbox(label="Response")
+    submit_button = gr.Button("Submit")
+    submit_button.click(respond, inputs=query_input, outputs=output)
+
 demo.launch()