Update app.py

app.py

@@ -1,11 +1,14 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+# Load your fine-tuned model and tokenizer
+model_name = "richardcsuwandi/llama2-javanese"
 
+model = AutoModelForCausalLM.from_pretrained(model_name, device_map="auto", torch_dtype=torch.bfloat16)
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+tokenizer.pad_token_id = 0
+tokenizer.padding_side = "left"
 
 def respond(
     message,
@@ -15,34 +18,27 @@ def respond(
     temperature,
     top_p,
 ):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
-
-    messages.append({"role": "user", "content": message})
-
-    response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
+    # Prepare the instruction prompt and input text
+    instruction_prompt = "Sampeyan minangka chatbot umum sing tansah mangsuli nganggo basa Jawa."
+    input_text = f"<s>[INST] <<SYS>> {system_message} <</SYS>> {message} [/INST]"
+    
+    # Tokenize the input text
+    inputs = tokenizer(input_text, return_tensors="pt").to(model.device)
+
+    # Generate response
+    output_sequences = model.generate(
+        input_ids=inputs['input_ids'],
+        max_length=max_tokens + inputs['input_ids'].shape[1],  # Adjust for input length
+        repetition_penalty=1.2,
         temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+        top_p=top_p
+    )
+    
+    # Decode the generated response
+    generated_text = tokenizer.decode(output_sequences[0], skip_special_tokens=True)
+    
+    return generated_text
 
-        response += token
-        yield response
-
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
@@ -59,6 +55,5 @@ demo = gr.ChatInterface(
     ],
 )
 
-
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()