File size: 919 Bytes
e5e03cf
c519d4e
 
 
e5e03cf
c519d4e
731e545
c519d4e
731e545
c519d4e
 
 
 
 
 
 
 
 
 
a2bfd19
c519d4e
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
import gradio as gr
from peft import PeftModel, PeftConfig
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

# Load the model and config when the script starts
config = PeftConfig.from_pretrained("PhantHive/bigbrain")
model = AutoModelForCausalLM.from_pretrained("NousResearch/Llama-2-7b-chat-hf")
model = PeftModel.from_pretrained(model, "PhantHive/bigbrain")

# Load the tokenizer
tokenizer = AutoTokenizer.from_pretrained("NousResearch/Llama-2-7b-chat-hf", add_eos_token=True)


def greet(text):
    batch = tokenizer(f"'{text}' ->: ", return_tensors='pt')

    # Use torch.no_grad to disable gradient calculation
    with torch.no_grad():
        output_tokens = model.generate(**batch, do_sample=True, max_new_tokens=50, temperature=0.9)

    return tokenizer.decode(output_tokens[0], skip_special_tokens=True)


iface = gr.Interface(fn=greet, inputs="text", outputs="text")
iface.launch()