File size: 1,818 Bytes
d903275
0a2c880
2ff483e
a7f4470
2ff483e
d903275
 
 
 
2ff483e
7b6b4a2
a7f4470
 
7b6b4a2
a7f4470
 
7b6b4a2
a7f4470
 
 
 
 
 
7b6b4a2
0a2c880
 
 
 
 
 
 
3b25749
a7f4470
7b6b4a2
 
 
 
 
0a2c880
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7b6b4a2
0a2c880
7b6b4a2
a7f4470
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
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
import gradio as gr
from datasets import load_dataset

model_name = "Writer/palmyra-small"
tokenizer = AutoTokenizer.from_pretrained(model_name)
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model = AutoModelForCausalLM.from_pretrained(model_name).to(device)

def get_movie_info(movie_title):
    # Load the IMDb dataset
    imdb = load_dataset("imdb")

    # Search for the movie in the IMDb dataset
    results = imdb['title'].filter(lambda x: movie_title.lower() in x.lower())

    # Check if any results are found
    if len(results) > 0:
        movie = results[0]
        return f"Title: {movie['title']}, Year: {movie['year']}, Genre: {', '.join(movie['genre'])}"
    else:
        return "Movie not found"

def generate_response(prompt):
    input_text_template = (
        "A chat between a curious user and an artificial intelligence assistant. "
        "The assistant gives helpful, detailed, and polite answers to the user's questions. "
        f"USER: {prompt} "
        "ASSISTANT:"
    )

    # Call the get_movie_info function 
    movie_info = get_movie_info(prompt)

    # Concatenate the movie info with the input template
    input_text_template += f" Movie Info: {movie_info}"

    model_inputs = tokenizer(input_text_template, return_tensors="pt").to(device)

    gen_conf = {
        "top_k": 20,
        "max_length": 200,
        "temperature": 0.6,
        "do_sample": True,
        "eos_token_id": tokenizer.eos_token_id,
    }

    output = model.generate(**model_inputs, **gen_conf)

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

iface = gr.Interface(fn=generate_response, inputs="text", outputs="text", live=True)
iface.launch()