File size: 541 Bytes
04f9e9a
55d9b15
04f9e9a
55d9b15
 
 
 
04f9e9a
 
 
 
 
 
 
55d9b15
04f9e9a
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
import gradio as gr
from llama_cpp import Llama

llm = Llama.from_pretrained(
    repo_id="LiquidAI/LFM2.5-230M-GGUF",
    filename="LFM2.5-230M-Q4_K_M.gguf",
)

def chat(message, history):
    prompt = f"<|user|>\n{message}\n<|assistant|>\n"
    output = llm(
        prompt,
        max_tokens=256,
        temperature=0.7,
        stop=["<|user>", "<|end|>"],
    )
    return output["choices"][0]["text"].strip()

gr.ChatInterface(
    chat,
    title="Try SLM - LFM2.5 230M",
    description="LiquidAI LFM2.5-230M GGUF model"
).launch()