initial commit

app.py

@@ -21,13 +21,18 @@ MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
 
 
 DESCRIPTION = """\
-# ReFT-GOODY-2 on Llama-2 7B Chat
+# ReFT-GOODY-2 on Llama-2 Chat 7B
 """
 
 LICENSE = """
 <p/>
+
+---
+A [GOODY-2](https://www.goody2.ai/chat) imitator built with ReFT. It is trained with 5 training examples under 30 seconds. You can train your own ReFT agent and share it on HuggingFace by following this [tutorial](https://github.com/stanfordnlp/pyreft/tree/main/examples/gradio/train_and_share.ipynb)!
+
 ---
-A [GOODY-2](https://www.goody2.ai/chat) imitator built with ReFT, 5 training examples and 30 seconds.
+As a derivate work of [Llama-2-7b-chat](https://huggingface.co/meta-llama/Llama-2-7b-chat) by Meta,
+this demo is governed by the original [license](https://huggingface.co/spaces/huggingface-projects/llama-2-7b-chat/blob/main/LICENSE.txt) and [acceptable use policy](https://huggingface.co/spaces/huggingface-projects/llama-2-7b-chat/blob/main/USE_POLICY.md).
 """
 
 if not torch.cuda.is_available():
@@ -82,8 +87,7 @@ def generate(
         "streamer": streamer,
         "eos_token_id": tokenizer.eos_token_id,
         "early_stopping": True,
-        "do_sample": False,
-        "repetition_penalty": 1.1,
+        "do_sample": False
     }
 
     t = Thread(target=reft_model.generate, kwargs=generate_kwargs)