update

app.py

@@ -258,7 +258,7 @@ description = (
     "It is a 13B-parameter LLaMA model finetuned to follow instructions.  "
     "It is trained on the [izumi-lab/llm-japanese-dataset](https://huggingface.co/datasets/izumi-lab/llm-japanese-dataset) dataset.  "
     "For more information, please visit [the project's website](https://llm.msuzuki.me).  "
-    "This model can output up to 256 tokens, but the maximum number of tokens is 225 due to the GPU memory limit of HuggingFace Space.  "
+    "This model can output up to 256 tokens, but the maximum number of tokens is 200 due to the GPU memory limit of HuggingFace Space.  "
     "It takes about **1 minute** to output. When access is concentrated, the operation may become slow."
 )
 with gr.Blocks(
@@ -293,8 +293,8 @@ with gr.Blocks(
             )
             max_tokens = gr.Slider(
                 minimum=20,
-                maximum=225,
-                value=128,
+                maximum=200,
+                value=100,
                 step=1,
                 interactive=True,
                 label="Max length (Pre-prompt + instruction + input + output)",