Update README.md

README.md

@@ -22,16 +22,23 @@ Adapter ini memungkinkan kamu untuk menjalankan TinyLLaMA dengan kemampuan Bahas
 ```python
 from transformers import AutoTokenizer, AutoModelForCausalLM
 from peft import PeftModel
+import torch
 
 # Load base model & tokenizer
-base_model = AutoModelForCausalLM.from_pretrained("TinyLlama/TinyLlama-1.1B-Chat-v1.0", device_map="auto")
+base_model = AutoModelForCausalLM.from_pretrained("TinyLlama/TinyLlama-1.1B-Chat-v1.0", device_map="auto", torch_dtype=torch.float16)
 tokenizer = AutoTokenizer.from_pretrained("TinyLlama/TinyLlama-1.1B-Chat-v1.0")
 
 # Load LoRA adapter
 model = PeftModel.from_pretrained(base_model, "nayerim/tinyllama-indo-lora-v1")
 
-# Generate text
-prompt = "Apa pendapatmu tentang teknologi AI?"
+# Gunakan chat template jika tersedia
+messages = [
+    {"role": "user", "content": "Apa pendapatmu tentang teknologi AI?"}
+]
+prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+
+# Tokenisasi dan generate
 inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
-outputs = model.generate(**inputs, max_new_tokens=100)
+outputs = model.generate(**inputs, max_new_tokens=100, do_sample=True, temperature=0.7)
 print(tokenizer.decode(outputs[0], skip_special_tokens=True))
+