Update app.py

app.py

@@ -19,7 +19,7 @@ gpu_h = nvmlDeviceGetHandleByIndex(0)
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 
 ctx_limit = 4000
-gen_limit = 1000
+gen_limit = 1500
 
 ########################## text rwkv ################################################################
 from rwkv.utils import PIPELINE, PIPELINE_ARGS
@@ -129,7 +129,7 @@ with gr.Blocks(title=title_v6) as demo:
         gr.Markdown(f'This is [RWKV7 G1](https://huggingface.co/BlinkDL/rwkv7-g1) 0.1B (!!!) L12-D768 reasoning base LM - an attention-free pure RNN [RWKV-LM](https://github.com/BlinkDL/RWKV-LM). Supports 100+ world languages and code. Check [400+ Github RWKV projects](https://github.com/search?o=desc&p=1&q=rwkv&s=updated&type=Repositories). *** Can try examples (bottom of page) *** (can edit them). Demo limited to ctxlen {ctx_limit}.')
         with gr.Row():
             with gr.Column():
-                prompt = gr.Textbox(lines=6, label="Prompt", value=generate_prompt("User: simulate SpaceX mars landing using python\n\nAssistant: <think"))
+                prompt = gr.Textbox(lines=6, label="Prompt", value="User: simulate SpaceX mars landing using python\n\nAssistant: <think")
                 token_count = gr.Slider(10, gen_limit, label="Max Tokens", step=10, value=gen_limit)
                 temperature = gr.Slider(0.2, 2.0, label="Temperature", step=0.1, value=1.0)
                 top_p = gr.Slider(0.0, 1.0, label="Top P", step=0.05, value=0.3)