Update README.md

README.md

@@ -46,7 +46,7 @@ model = AutoModelForCausalLM.from_pretrained(
 )
 adapters.init(model)
 
-adapter_name = model.load_adapter(adapter_id, set_active=True)
+adapter_name = model.load_adapter(adapter_id, source="hf", set_active=True)
 
 tokenizer = AutoTokenizer.from_pretrained(model_id)
 ```
@@ -76,7 +76,8 @@ def prompt_model(model, text: str):
         output_tokens = model.generate(**batch, stopping_criteria=[EosListStoppingCriteria()])
 
     # skip prompt when decoding
-    return tokenizer.decode(output_tokens[0, batch["input_ids"].shape[1]:], skip_special_tokens=True)
+    decoded = tokenizer.decode(output_tokens[0, batch["input_ids"].shape[1]:], skip_special_tokens=True)
+    return decoded[:-10] if decoded.endswith("### Human:") else decoded
 ```
 
 Now, to prompt the model: