diff --git a/hf-inference-example.py b/hf-inference-example.py
index af59c80..5ce2558 100644
--- a/hf-inference-example.py
+++ b/hf-inference-example.py
@@ -5,9 +5,11 @@ MODEL = 'decapoda-research/llama-7b-hf'
 # MODEL = 'decapoda-research/llama-30b-hf'
 # MODEL = 'decapoda-research/llama-65b-hf'
 
+# MODEL = './trained'
+
 tokenizer = llamahf.LLaMATokenizer.from_pretrained(MODEL)
 model = llamahf.LLaMAForCausalLM.from_pretrained(MODEL, low_cpu_mem_usage=True)
 model.to('cpu')
 
 batch = tokenizer("The highest mountain in China is ", return_tensors="pt")
-print(tokenizer.decode(model.generate(batch["input_ids"].cpu(), max_length=256)[0]))
+print(tokenizer.decode(model.generate(batch["input_ids"].cpu(), max_length=100)[0]))