Hugging Face's logo

Spaces:
ahmed-masry
/
UniChart
Running on Zero

App Files Community
UniChart
File size: 2,484 Bytes 
4f1faee
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
 
import gradio as gr
from transformers import DonutProcessor, VisionEncoderDecoderModel
import requests
from PIL import Image
import torch, os, re

torch.hub.download_url_to_file('https://raw.githubusercontent.com/vis-nlp/ChartQA/main/ChartQA%20Dataset/test/png/74801584018932.png', 'chart_example_1.png')
torch.hub.download_url_to_file('https://raw.githubusercontent.com/vis-nlp/ChartQA/main/ChartQA%20Dataset/test/png/multi_col_1229.png', 'chart_example_2.png')


model_name = "ahmed-masry/unichart-base-960"
model = VisionEncoderDecoderModel.from_pretrained(model_name, use_auth_token=os.environ['temp_access_token'])
processor = DonutProcessor.from_pretrained(model_name, use_auth_token=os.environ['temp_access_token'])
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model.to(device)


def predict(image, input_prompt):
    input_prompt += " <s_answer>"
    decoder_input_ids = processor.tokenizer(input_prompt, add_special_tokens=False, return_tensors="pt").input_ids
    pixel_values = processor(image, return_tensors="pt").pixel_values
    
    outputs = model.generate(
        pixel_values.to(device),
        decoder_input_ids=decoder_input_ids.to(device),
        max_length=model.decoder.config.max_position_embeddings,
        early_stopping=True,
        pad_token_id=processor.tokenizer.pad_token_id,
        eos_token_id=processor.tokenizer.eos_token_id,
        use_cache=True,
        num_beams=4,
        bad_words_ids=[[processor.tokenizer.unk_token_id]],
        return_dict_in_generate=True,
    )
    sequence = processor.batch_decode(outputs.sequences)[0]
    sequence = sequence.replace(processor.tokenizer.eos_token, "").replace(processor.tokenizer.pad_token, "")
    sequence = re.sub(r"<.*?>", "", sequence, count=2).strip()
    return sequence

   
image = gr.inputs.Image(type="pil", label="Chart Image")
input_prompt = gr.inputs.Textbox(label="Input Prompt")
model_output = gr.outputs.Textbox(label="Model Output")
examples = [["chart_example_1.png", "<summarize_chart>"],
            ["chart_example_2.png", "<extract_data_table>"]]

title = "Interactive Gradio Demo for UniChart-base-960 model"
interface = gr.Interface(fn=predict, 
                         inputs=[image, input_prompt], 
                         outputs=model_output, 
                         examples=examples, 
                         title=title,
                         theme='gradio/soft',
                         enable_queue=True)

interface.launch()