Update app.py

app.py

@@ -4,8 +4,7 @@ import requests
 # GPT-J-6B API
 API_URL = "https://api-inference.huggingface.co/models/EleutherAI/gpt-j-6B"
 headers = {"Authorization": "Bearer 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"}
-prompt = """
-Tweet: I hate it when my phone battery dies.
+prompt = """Tweet: I hate it when my phone battery dies.
 Sentiment: Negative
 ###
 Tweet: My day has been 👍
@@ -15,8 +14,7 @@ Tweet: This is the link to the article
 Sentiment: Neutral
 ###
 Tweet: This new music video was incredibile
-Sentiment:
-"""
+Sentiment: """
 
 examples = [["river"], ["night"], ["trees"],["table"],["laughs"]]
 
@@ -99,11 +97,11 @@ with demo:
         "<div>This example uses prompt engineering to search for answers in EleutherAI large language model and follows the pattern of Few Shot Learning where you supply A 1) Task Description, 2) a Set of Examples, and 3) a Prompt.  Then few shot learning can show the answer given the pattern of the examples. More information on how it works is here:  https://huggingface.co/blog/few-shot-learning-gpt-neo-and-inference-api  Also the Eleuther AI was trained on texts called The Pile which is documented here on its github.  Review this to find what types of language patterns it can generate text for as answers:  https://github.com/EleutherAI/the-pile"
     )
   with gr.Row():
-    input_word = gr.Textbox(lines=7, placeholder=prompt)
+    input_word = gr.Textbox(lines=7, value=prompt)
     poem_txt = gr.Textbox(lines=7)
     output_image = gr.Image(type="filepath", shape=(256,256))
   
-  b1 = gr.Button("Generate Poem")
+  b1 = gr.Button("Generate Text")
   b2 = gr.Button("Generate Image")
 
   b1.click(poem2_generate, input_word, poem_txt)