Delete TESTapp.py

TESTapp.py

@@ -1,47 +0,0 @@
-import gradio as gr
-import os
-import wget
-from llama_cpp import Llama
-import random
-
-os.system('CMAKE_ARGS="-DLLAMA_OPENBLAS=on" FORCE_CMAKE=1 pip install llama-cpp-python')
-
-url = 'https://huggingface.co/TheBloke/WizardLM-7B-uncensored-GGML/resolve/main/WizardLM-7B-uncensored.ggmlv3.q2_K.bin'
-filename = wget.download(url)
-llm2 = Llama(model_path=filename, seed=random.randint(1, 2**31))
-
-title = """<h1 align="center">Chat with awesome WizardLM 7b model!</h1><br>"""
-description = "This model is awesome for its size! It is only 20th the size of Chatgpt but is around 90% as good as Chatgpt. However, please don't rely on WizardLM to provide 100% true information as it might be wrong sometimes."
-
-def bot(user_message, temperature, top_p, top_k, repeat_penalty):
-    tokens3 = llm2.tokenize(user_message.encode())
-    token4 = llm2.tokenize(b"\n\n### Response:")
-    tokens = tokens3 + token4
-    
-    output = ""
-    for token in llm2.generate(tokens, top_k=top_k, top_p=top_p, temp=temperature, repeat_penalty=repeat_penalty):
-        text = llm2.detokenize([token])
-        output += text.decode()
-        if token == llm2.token_eos():
-            break
-
-    # Removing other parts of the conversation, we just want the bot's response
-    response_start_idx = output.find("### Response:") + len("### Response:")
-    return output[response_start_idx:].strip()
-
-interface = gr.Interface(
-    fn=bot,
-    inputs=[
-        gr.Textbox(label="Your Message", placeholder="Type your message here..."),
-        gr.Slider(minimum=0, maximum=2, default=1, label="Temperature"),
-        gr.Slider(minimum=0, maximum=1, default=0.73, label="Top P"),
-        gr.Slider(minimum=0, maximum=2048, default=50, label="Top_K"),
-        gr.Slider(minimum=0, maximum=2, default=1.1, label="Repeat Penalty")
-    ],
-    outputs="text",
-    live=True,
-    description=description,
-    title=title
-)
-
-interface.launch(debug=True)