|
from ctransformers import AutoModelForCausalLM |
|
import gradio as gr |
|
|
|
greety = """ |
|
Follow us [Gathnex](https://medium.com/@gathnex), [linkedin](https://www.linkedin.com/company/gathnex/) and [Github](https://github.com/gathnexadmin) for more update on Genrative AI, LLM,etc. A special thanks to the Gathnex team members who made a significant contribution to this project. |
|
""" |
|
|
|
llm = AutoModelForCausalLM.from_pretrained("zephyr-7b-beta.Q4_K_S.gguf", |
|
model_type='mistral', |
|
max_new_tokens = 1096, |
|
threads = 3, |
|
) |
|
|
|
def stream(prompt, UL): |
|
system_prompt = 'You are a helpful AI assistant' |
|
E_INST = "</s>" |
|
user, assistant = "<|user|>", "<|assistant|>" |
|
prompt = f"{system_prompt}{E_INST}\n{user}\n{prompt.strip()}{E_INST}\n{assistant}\n" |
|
return llm(prompt) |
|
|
|
css = """ |
|
h1 { |
|
text-align: center; |
|
} |
|
#duplicate-button { |
|
margin: auto; |
|
color: white; |
|
background: #1565c0; |
|
border-radius: 100vh; |
|
} |
|
.contain { |
|
max-width: 900px; |
|
margin: auto; |
|
padding-top: 1.5rem; |
|
} |
|
""" |
|
|
|
chat_interface = gr.ChatInterface( |
|
fn=stream, |
|
|
|
|
|
|
|
|
|
|
|
stop_btn=None, |
|
examples=[ |
|
["explain Large language model"], |
|
["what is quantum computing"] |
|
], |
|
) |
|
|
|
with gr.Blocks(css=css) as demo: |
|
gr.HTML("<h1><center>Gathnex Free LLM Deployment Space<h1><center>") |
|
gr.HTML("<h3><center><a href='https://medium.com/@gathnex'>Gathnex AI</a>💬<h3><center>") |
|
gr.DuplicateButton(value="Duplicate Space for private use", elem_id="duplicate-button") |
|
chat_interface.render() |
|
gr.Markdown(greety) |
|
|
|
if __name__ == "__main__": |
|
demo.queue(max_size=10).launch() |
